Deaerator Tank Assessment Guideline PDF
Deaerator Tank Assessment Guideline PDF
Deaerator Tank Assessment Guideline PDF
1021199
11179101
11179101
Deaerator Tank Assessment Guideline
Current Industry Technology and Approaches
1021199
11179101
DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES
THIS DOCUMENT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN ACCOUNT OF
WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI).
NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE ORGANIZATION(S) BELOW, NOR ANY
PERSON ACTING ON BEHALF OF ANY OF THEM:
(A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH
RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM
DISCLOSED IN THIS DOCUMENT, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED
RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL PROPERTY, OR (III) THAT THIS DOCUMENT IS
SUITABLE TO ANY PARTICULAR USER'S CIRCUMSTANCE; OR
(B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING
ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED
OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS
DOCUMENT OR ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN
THIS DOCUMENT.
This is an EPRI Technical Update report. A Technical Update report is intended as an informal report of
continuing research, a meeting, or a topical study. It is not a final EPRI technical report.
NOTE
For further information about EPRI, call the EPRI Customer Assistance Center at 800.313.3774 or
e-mail [email protected].
Electric Power Research Institute, EPRI, and TOGETHER…SHAPING THE FUTURE OF ELECTRICITY
are registered service marks of the Electric Power Research Institute, Inc.
Copyright © 2010 Electric Power Research Institute, Inc. All rights reserved.
11179101
ACKNOWLEDGMENTS
The following organization, under contract to the Electric Power Research Institute (EPRI),
prepared this report:
Principal Investigators
M. Moskal
C. Noble
Contributing Engineers
R. Munson
D. Daniels
G. M. Tanner
B. Bruscato
EPRI would like to acknowledge the contributions of the following utilities and personnel for
providing technical guidance, pictures/examples of damage, and case studies:
This publication is a corporate document that should be cited in the literature in the following
manner:
Deaerator Tank Assessment Guideline: Current Industry Technology and Approaches. EPRI,
Palo Alto, CA: 2010. 1021199.
11179101
Todd Kuntz Arizona Public Service (APS)
Ecodyne Limited
iv
11179101
v
11179101
11179101
ABSTRACT
The leaking and rupture of deaerator (DA) vessels has been an issue in industrial plants since the
early 1970s. Three catastrophic failures of DA vessels that occurred in North American plants in
1983 prompted an industry-wide effort to understand and prevent these failures. The National
Association of Corrosion Engineers (NACE) first published their Standard Practice: Prevention,
Detection, and Correction of Deaerator Cracking SP0590 in April 1990 (the latest version is
2007). Inspection of DA systems across all industry segments has identified high percentages of
these vessels with cracking (greater than 30%). The most recent NACE document reports 38% of
the vessels inspected at utilities have cracks requiring repair. Unfortunately, unit cycling and
lower maintenance frequencies dictated by current economic conditions will likely intensify the
severity of the DA cracking problems.
The goal of this project was to revise and update the Electric Power Research Institute (EPRI)
report Deaerator Vessel Assessment Guideline (DVAG) to reflect current industry technology
and approaches. This updated guideline incorporates technical input from the NACE Standard
Practice Document SP0590-2007, but is focused on utility applications. This document provides
a comprehensive guideline of today’s “best practices,” including information and guidance on
the following:
• Design and materials of construction.
• Advanced understanding of damage: description and explanation of damage mechanisms,
including examples of failure.
• Operational conditions that promote failure.
• Inspection and evaluation.
• Repair options.
• Technical criteria for the mitigation of damage including recommendations of repair-run-or-
replace. This includes references to current fitness-for-service best practices such as ASME
FFS-001 and API 579.
Overall, this document allows plant engineers, plant chemists, and operators/owners to identify
the risk in their DA systems and provides sound engineering evaluation practices for running
vessels with damage, repairing them if necessary, or replacing them with vessels with greater
integrity.
vii
11179101
Keywords
Corrosion fatigue
Deaerator vessel
Fitness-for-service
Flow-accelerated corrosion
Pressure vessel
Steam impingement
viii
11179101
ACRONYMS
CF – corrosion fatigue
DA – deaerator
ET – electromagnetic testing
FFS – fitness-for-service
LP – low pressure
ix
11179101
MT – magnetic particle testing
OT – oxygenated treatment
QC – quality control
UT – ultrasonic testing
VT – visual inspection
x
11179101
CONTENTS
1 INTRODUCTION/BACKGROUND........................................................................................1-1
1.1 Principals of Operation ..................................................................................................1-1
1.2 Types of Deaerators ......................................................................................................1-1
1.3 EPRI Licensed Material .................................................................................................1-2
1.3.1 Design Basis ..........................................................................................................1-2
1.3.2 Components of Deaerators ....................................................................................1-2
1.3.2.1 General ..........................................................................................................1-2
1.3.2.2 Internals .........................................................................................................1-3
1.4 Current Trends ..............................................................................................................1-5
1.5 EPRI Licensed Material .................................................................................................1-6
1.5.1 Failure History ........................................................................................................1-6
1.5.2 New Construction Recommendations ....................................................................1-6
1.6 References ....................................................................................................................1-7
1.6.1 References Cited in Text ........................................................................................1-7
1.6.2 Other Sources ........................................................................................................1-7
xi
11179101
2.8.1 References Cited in Text ......................................................................................2-12
2.8.2 Other Sources ......................................................................................................2-12
xii
11179101
4.1.4 CF Remedies .........................................................................................................4-5
4.2 Flow-Accelerated Corrosion (FAC) ................................................................................4-5
4.2.1 General ..................................................................................................................4-5
4.2.2 Description .............................................................................................................4-6
4.2.3 Where FAC is Found .............................................................................................4-6
4.2.4 FAC Remedies ......................................................................................................4-6
4.3 Corrosion Pitting (Downtime Corrosion).........................................................................4-7
4.3.1 General ..................................................................................................................4-7
4.3.2 Where Pitting Corrosion is Found ..........................................................................4-7
4.3.3 Pitting Remedies ....................................................................................................4-7
4.4 Steam Impingement ......................................................................................................4-8
4.4.1 General ..................................................................................................................4-8
4.4.2 Where Steam Impingement is Found .....................................................................4-8
4.4.3 Steam Impingement Remedies ..............................................................................4-8
4.5 Mechanical Damage ......................................................................................................4-8
4.5.1 General ..................................................................................................................4-8
4.5.2 Identification...........................................................................................................4-9
4.5.3 Mechanical Damage Remedies .............................................................................4-9
4.6 Less Common Damage Mechanisms ............................................................................4-9
4.6.1 General ..................................................................................................................4-9
4.6.2 Stress Corrosion Cracking (SCC) ..........................................................................4-9
4.6.3 Thermal Fatigue Cracking ....................................................................................4-10
4.6.4 Down Time Corrosion ..........................................................................................4-10
4.6.5 Corrosion Grooving ..............................................................................................4-10
4.6.6 Mechanical Fatigue (Vibration Fatigue) ................................................................4-10
4.6.7 Galling .................................................................................................................4-11
4.6.8 Preferential Corrosion ..........................................................................................4-11
4.6.9 Other....................................................................................................................4-11
4.7 References ..................................................................................................................4-12
4.7.1 References Cited in Text ......................................................................................4-12
4.7.2 Other Sources ......................................................................................................4-13
xiii
11179101
5.3 Feedwater Chemistry and Flow-Accelerated Corrosion .................................................5-3
5.4 Feedwater Chemistry and Corrosion Fatigue (Stress Assisted Cracking) ......................5-5
5.5 EPRI References for Water Chemistry ..........................................................................5-8
5.5.1 EPRI Documents on FAC ......................................................................................5-8
5.5.2 EPRI Documents on Corrosion Fatigue .................................................................5-8
5.5.3 EPRI Chemistry Guidelines for Fossil Fired Power Plants ......................................5-9
xiv
11179101
8.2.1 Brittle Failures ........................................................................................................8-2
8.2.2 General Metal Loss ................................................................................................8-2
8.2.3 Localized Metal Loss .............................................................................................8-2
8.2.4 Pitting Corrosion ....................................................................................................8-3
8.2.5 Shell Distortion .......................................................................................................8-3
8.2.6 Crack-Like Flaws ...................................................................................................8-3
8.2.7 Dents and Gouges .................................................................................................8-4
xv
11179101
11179101
LIST OF FIGURES
xvii
11179101
Figure 10-4 Flow accelerated corrosion of the DA heater circumferential weld seam
revealed a partial penetration weld. The DA heater was replaced. .................................10-7
Figure 10-5 Drawing shows the failure location in the deaerator heater .................................10-8
xviii
11179101
LIST OF TABLES
xix
11179101
11179101
Introduction/Background
1
INTRODUCTION/BACKGROUND
Deaerators are an essential piece of equipment in most sub-critical boilers and are used for
oxygen removal as well as storage of boiler feedwater. This section gives background
information on the operation and types of deaerators. 1
In a general sense, a deaerator is nothing more than a direct contact feedwater heater with the
added benefit of removing non-condensable gases (NCGs). To accomplish this, the inlet water
(makeup and/or condensate) is mixed with the heating steam. Since the heating steam is assumed
to be free from NCGs, the gases move from the water to the steam space because the partial
pressure of the NCGs in the steam space is essentially zero. A portion of the steam and the NCGs
are then vented out of the unit (typically to atmosphere) to maintain the low partial pressure of
the NCGs.
Note that for the most efficient transfer of the NCGs from the inlet water to the steam space, the
water is heated to the saturation temperature corresponding to the operating pressure. In addition,
the surface area of the water is maximized by spraying it into fine droplets and using mechanical
means to continuously break up the water.
Finally, there is a certain amount of residence time during which the water droplets must be
exposed to the NCG-free steam so that the NCGs can “migrate” out of the water. This residence
time is a function of the inlet temperature, operating pressure, and droplet size (as determined by
the type of internals in the deaerator).
There are essentially two types of deaerators in use today—spray (or spray-scrubber) units and
tray (or spray-tray) units. Spray units utilize an atomizing spray section to break up the water for
initial heating and deaeration. This spray stage is then followed with a scrubber section which
uses fixed baffles and/or orifices to promote vigorous mixing of the heating steam and inlet
water to scrub out the remaining traces of NCGs.
1
This introductory section is repeated (with a few edits) from the EPRI Repair of Deaerators document, ID
#1008069, June 2004.
1-1
11179101
Introduction/Background
Tray units utilize a similar spray section for initial heating and deaeration, but follow this with a
tray section (in place of the scrubber) which cascades the inlet water over a series of trays while
heating steam flows through the tray stack to scrub out the remaining NCGs.
Both types of units are capable of removing oxygen down to 7 ppb or less, but the tray design
has better performance as loads vary, and typically a longer life due to containment of NCGs
within a stainless steel tray enclosure. The vast majority of deaerators in utility power plants are
tray units and the information in this document is biased toward them; however many of the
principles also apply to spray units.
Most deaerators in the US are designed and constructed to ASME Section VIII, Division 1. This
Code explains how to calculate minimum wall thickness of pressure vessels for a given operating
temperature and pressure. The minimum wall thickness accounts for the extra wall thickness
needed for corrosion allowance. It also sets forth requirements for construction including
materials, inspection, and fabrication, which includes welding, preheating, and post weld heat
treatment (PWHT). Section 3 of this document covers Code details for deaerators.
1.3.2.1 General
Most utility deaerators in the United States are two tank designs, though some single tank
designs are used domestically. Single tank designs are common overseas. The typical two-tank
design (see Figure 1-1) consists of a deaerating heater vessel, where the steam and water are
mixed, and a storage tank. The storage tank serves several purposes:
2. To provide reserve and measurement of the NPSH available to the boiler feedwater pumps.
3. To provide reaction time for water treatment chemical injected into the feedwater
Although the storage tank is typically just a large tank, its duty can be severe and it should be
subjected to the same rigorous inspection as the heater vessel. In fact, numerous deaerator
failures have been due to problems with the storage tank.
1-2
11179101
Introduction/Background
Figure 1-1
Deaerator in operation
1.3.2.2 Internals
The remaining components are known collectively as the internals (see Figure 1-2 for typical
arrangement.). Water enters the unit through a spray header or waterbox, is sprayed through
spray valves, and cascades over the trays which are contained in the tray enclosure (see
Figure 1-3 and Figure 1-4 for tray and spray valve photographs.). The primary purpose of the
internals is to provide sufficient contact and residence time to heat and deaerate the feedwater. In
addition, the internals serve to contain the non-condensable gases and prevent them from
attacking the pressure shell of the unit.
Although outdated, external vent condensers are occasionally used to minimize the vent plume
and associated steam loss from the unit. These are typically an external shell and tube heat
exchanger utilizing the inlet makeup stream to condense a significant amount of the vent steam.
The drains from the exchanger are then gravity fed back to the heating area over the trays. Most
modern deaerators feature an internal vent condenser to minimize the required vent rate. This is
an area inside the unit where the vent stream passes in close proximity to the inlet header or
waterbox. The cooling effect of the inlet water serves to condense some of the vent stream for a
lesser overall vent rate.
1-3
11179101
Introduction/Background
Figure 1-2
Deaerator internals
Figure 1-3
Deaerator trays
1-4
11179101
Introduction/Background
Figure 1-4
Deaerator spray valve
Deaerators have come under increasing scrutiny since the late 1980s when there were a few
instances of the pressure vessel failing catastrophically with some loss of life. The failures were
examined by NACE and their findings formed the basis of NACE Standard RP0590-90 (the
current version is SP0590-2007) [1]. The final result for most plants is that their insurance carrier
requires more frequent and thorough inspection of the deaerator to evaluate its integrity as a
pressure vessel.
Deaerators have also been included in the recent research into flow-accelerated corrosion (FAC).
While originally researched as a pipeline issue in nuclear plants, FAC has been observed in
numerous deaerators and has caused at least one through-wall leak in a utility deaerator. FAC
will be discussed in more detail in Section 4 of this document titled “Damage Mechanisms.”
Finally, the utility of deaeration as a whole has come into question as some utility plants have
switched to oxygenated treatment. Oxygenated treatment has been in extensive use overseas for
some years but has only recently come into play in the U.S. Essentially, plants with all ferrous
feedwater systems have had some success in allowing dissolved oxygen levels to run
considerably higher than traditional all-volatile treatments allow. The intent is to form an oxide
layer on the feedwater components which protects them from further attacks.
Oxygenated treatment will reduce the amount of oxygen removal that a deaerator must achieve,
but it will not make the deaerator obsolete. Even base loaded plants will need a deaerator for
start-up use and obviously any existing plant will still need the thermal duty of the deaerator as a
feedwater heater. In addition, many of the plants in the U.S. do not have the appropriate
metallurgy to use oxygenated treatment and will have to continue to rely upon the deaerator to
minimize oxygen attack.
1-5
11179101
Introduction/Background
In 1984, a NACE task group was formed to study the high incidence of cracking problems in
deaerators. The result of this study is the aforementioned NACE Standard RP0590-90.
Considered in this study were three storage vessel ruptures, one of which resulted in fatalities.
Also noted is a TAPPI “Deaerator Advisory” that reports cracks found in approximately 50% of
the vessels inspected in 1983. The noted cracking occurred in welds and their heat-affected
zones.
The NACE standard notes data that out of more than 700 vessels inspected, 30 to 40% contained
cracks and needed repairs. The cracking, found primarily through wet fluorescent magnetic
particle testing, is an environmentally assisted form described as corrosion fatigue that can occur
in either the deaerator or storage tank. The standard also reports that cracking is not a function of
vessel age. Cracking was found in the vapor zone of the deaerator, at the liquid/vapor interface in
the storage tank, and in the liquid zone of the storage tank. The NACE standard concludes that,
for horizontal vessels, the most prevalent area for cracking is in the liquid zone between the 4
o’clock and 8 o’clock positions.
It is important to note that one of the factors contributing to crack growth is residual tensile stress
from welding. As discussed in the design section of this document (Section 2), all deaerators and
deaerator storage tanks should be post weld heat treated to minimize residual stresses. It should
also be noted that PWHT does not eliminate the tendency towards fatigue cracking; some vessels
with PWHT have been found to be cracked.
As a result of deaerator failure studies by TAPPI and NACE, HEI and NACE have published
guidelines and recommendations for the construction of new deaerators, intended to reduce the
incidence of cracking and subsequent failure [1, 2]. The following is a summary of the
recommendations. Where the two standards differ, the more conservative of the two is given.
• Corrosion allowance – Vessel head and shell components are to be designed to include
1/8-inch corrosion allowance. Nozzles are to be designed to include 1/16-inch corrosion
allowance.
• Welding – Pressure retaining welds are to be full penetration and full fusion.
• Weld seam profile – Internal and external weld seams are to be smooth, and free from abrupt
changes. Welds are to blend ground as required.
• NDE of nozzle welds – Internal nozzle-to-shell welds are to be MT or PT examined as
applicable.
• Radiography – Shell and head seams are to be inspected by radiography to obtain a 1.0 joint
efficiency (i.e., RT-2 as per ASME Section VIII, Div. 1.)
1-6
11179101
Introduction/Background
• Post weld heat treatment (PWHT) – Vessels are to receive PWHT in accordance with ASME
code (normally 1100°F minimum for one hour per inch thickness).
• Trays – Trays are to be stamped or riveted, stainless steel construction.
• Materials – Materials are to be such that non-deaerated (or partially deaerated) water and
NCGs do not come into contact with any carbon steel vessel components.
1.6 References
1-7
11179101
11179101
2
DA CONSTRUCTION, MECHANICAL DESIGN, AND
MATERIALS
Deaerators remove dissolved oxygen and carbon dioxide by thermo-mechanical means. Under
pressure or vacuum, the water temperature is raised with steam, reducing the solubility of the
dissolved gases. Dissolved gases are removed from the deaerator by venting with steam. There is
a time associated with this reaction so residence time or dwell time is an important consideration.
Typical deaerator performance standards require that the temperature of the influent water shall
be within 2ºF (1.1ºC) of the saturation temperature at the specific operating pressure of the
deaerator, and to achieve 0.005 mL/L (7 PPB) outlet dissolved oxygen content. The design of
spray-tray type deaerators normally requires the above performance while operating from 10% to
100% of design capacity, which are typical requirements for utility applications.
2.2 Design
Experience shows that deaerator design has evolved to meet changing output performance
requirements and to provide equipment that will run year after year with little maintenance.
Deaerator equipment is available from many suppliers, though only a few typically supply the
needs for electric power utilities in North America. The deaerator designer is challenged to:
(1) achieve the required deaeration performance and output, typically specified as pounds per
hour (PPH) of water; (2) fit the equipment within the available space, sometimes a problem when
replacing existing equipment; (3) address known damage mechanisms of deaerators; (4) address
non-steady state operations such as turbine trips, cold water inlet surges, and deaerator pressure
loss; and (5) provide the equipment at a competitive price.
For example, deaerators with large diameter to length ratios provide the needed capacity and
water residence time for sufficient deaeration, but these larger vessels require heavier wall
thickness to meet pressure vessel code standards. Conversely, deaerators with small diameter to
length ratios, while lower in cost to manufacture, may be problematic with excessive steam flow
rates, sometimes leading to rapid vessel damage from flow-assisted corrosion (see Section 4 for
details.). Large storage tanks, up to 150 ft. (45.7 m) in length, can be disproportionally costly to
fabricate, require very large furnaces for post weld heat treatment and sometimes require
extensive field weld assembly. These are but a few of the challenges to producers of deaerator
equipment.
2-1
11179101
DA Construction, Mechanical Design, and Materials
Deaerator systems for power generation utilities are designed and built as either spray-tray
(sometimes referred to as simply “tray”) or spray-scrubber units. Although both systems can
achieve the above performance requirements, utility systems require the high capacity provided
by spray-tray deaerators. Spray-tray deaerators may be designed with varying orientations with
respect to the storage tank (Figure 2-1). In high capacity utility systems it is most common to
have a horizontal or vertical water deaerator vessel positioned over the storage tank, connected
by a short length of water downcomer and equalizing piping. A typical vertical deaerator and
storage tank with connections and names of parts is shown in Figure 2-2.
Figure 2-1
Orientation and combinations of deaerators and storage tanks. Source: Heat Exchange
Institute standards and typical specifications for tray-type deaerators.
2-2
11179101
DA Construction, Mechanical Design, and Materials
Figure 2-2
Typical vertical deaerator and storage tank with names of components and accessories.
Source: Heat Exchange Institute standards and typical specifications for tray-type
deaerators.
The components of a typical horizontal spray-tray system are shown in Figure 2-3. During
operation, most of the dissolved oxygen is released in the spray chamber when droplets contact
the non-condensable, free steam, but final oxygen is stripped when water droplets cascade
though the series of trays at the bottom of the deaerator vessel. The direction of steam flow
through the deaerator is typically countercurrent from the water direction (from the bottom up
2-3
11179101
DA Construction, Mechanical Design, and Materials
through the trays), but designs may be concurrent or even cross-current with water flow. The
major components within the tray enclosure are the water distribution system, the tray supports
and the trays.
Figure 2-3
Typical arrangement of horizontal deaerator with a waterbox (Courtesy of Kansas City
Deaerator Company, Inc.)
2-4
11179101
DA Construction, Mechanical Design, and Materials
Figure 2-4
Horizontal deaerator with a header pipe distribution system (Courtesy of Kansas City
Deaerator Company, Inc.)
2-5
11179101
DA Construction, Mechanical Design, and Materials
Water distribution systems, fabricated entirely from austenitic stainless steel, have improved in
the past two decades in attempts to avoid cracking problems associated with differential thermal
expansion. Most horizontal deaerators built prior to the 1990’s were designed with a “football”
shaped waterbox (Figure 2-3). The upper portion of the waterbox is comprised of sheet stainless
steel lining welded to the carbon steel shell; the lower half is fabricated from stainless steel plate
and contains the water spray valves. Depending on the particular design, this type of waterbox
arrangement may be prone to cracking at the long weld seam between the box and the shell.
Rapid temperature excursions also can result in distortion and cracking in the waterbox at the
vents and the spray valve locations. Later designs, shown in Figure 2-4, with the water pipe
header separated from the deaerator shell avoid many of the thermal expansion problems
associated with the contiguously welded stainless steel box and carbon steel shell.
At the top of the deaerator, non-condensable gases and steam collect and are vented through or
above the water distribution system. Because the environment is very corrosive, vents and vent
condensers must necessarily be made from stainless steel.
The tray stack, housed below the waterbox or pipe header, is supported by and contained within
the tray enclosure. The enclosure and its supports must be sturdily designed to hold the weight of
the tray stack. Hold-down clips and attachments must also be sufficiently rugged to restrain the
trays should an upset from water flashing occur in the storage section. All of the above
components, water distribution system, tray enclosure, and trays are typically fabricated from
stainless steel to resist corrosion from oxygenated water. (Industrial deaerators often bring
process condensate and makeup water directly into the deaerator. These streams are often
saturated with oxygen, and therefore, stainless steel components are required. Deaerator
manufacturers design for these more rugged conditions.)
In a typical utility boiler, the turbine condensate is pumped forward from the hotwell to the DA
with very little potential for dissolved oxygen to contaminate the feedwater. The current EPRI
guideline for dissolved oxygen at the condensate pump discharge, 10 ppb, is far too low to cause
oxygen pitting. For units that cycle frequently, particularly from a cold start, the stainless steel
spray headers and trays are still important as they will often see high dissolved oxygen levels
until the boiler is generating sufficient steam to provide steam to the deaerator.
Following the cascade through the trays, the fully deaerated water collects in the sump portion of
the deaerator vessel and proceeds through a vortex breaker and downcomer piping to the storage
tank. One or more equalizing pipe connections between the deaerator and the storage tank are
required to maintain equal pressure between the vessels.
The storage tank is simply a large tank to provide a stable net positive suction head for the
feedwater pumps. It also stores water and serves as a surge capacity for varying plant loads.
Finally, a system is provided for feedwater chemical additions, typically reducing agents such as
hydrazine, which are typically added downstream of the condensate pump discharge. However,
the storage tank is the first area in the feedwater system with sufficient temperature and
residence time to complete the chemical reaction.
2-6
11179101
DA Construction, Mechanical Design, and Materials
Finally, the deaerator and storage tank should be designed for good internal accessibility which
is needed for proper visual and nondestructive test inspections. Not all of the deaerator vessel can
normally be accessed for inspection, but consideration for needed inspections of known damage
mechanisms is vital to safe and long-term operation of the equipment.
Deaerator materials have not changed much for several decades, with pressure vessel grade
carbon steel being used for the shell, heads, and nozzles. Austenitic stainless steel has invariably
been specified for all components contacting un-deaerated or partially deaerated water: inlet
nozzle, waterbox/ header pipe, waterbox lining, vent condenser, valve plate, vent and spray
valves. Tray enclosures and trays are also specified to be stainless steel. A summary of current
materials of construction is shown in Table 2.1.
2-7
11179101
DA Construction, Mechanical Design, and Materials
Table 2-1
Materials of construction for deaerators
Strength (ksi)
Component Subcomponent Material Yield, Comments
Tensile
min.
Withdrawn in 1967; replaced
by SA515, Gr. 70 (coarse-
SA 212, Grade B 70-90 38
grained practice) and SA516,
Gr. 70 (fine-grained practice).
Intermediate strength steel.
SA 285, Grade B 50-70 27 May be ordered as copper-
bearing.
Shell and Heads
Intermediate strength steel.
SA 285, Grade C 55-75 30 May be ordered as copper-
bearing.
Deaerator
and Storage Steelmaking practice is coarse
Vessels SA 515, Grade 70 70-90 38 austenitic grain unless fine
grain is specified.
SA 516, Grade 70 70-90 38 Fine grained practice.
2-8
11179101
DA Construction, Mechanical Design, and Materials
Knowledge of the shell materials used in deaerators is important when repairs are contemplated
or required. Welding repair procedures vary for different carbon and low-alloy steels. Further,
the strength and toughness properties are necessary when Fitness-For-Service (FFS) assessments
must be made on damaged equipment.
Historically, almost all deaerator shells have been fabricated using one of several available
carbon steels. The specific grade of carbon steel plate varies according to grades and steel mill
practices at the time of manufacture. Deaerators manufactured prior to 1967 were largely
fabricated from SA 212, Grade B, which was superseded by SA 285, Grades B and C. About
1970, SA 515, Grade 70 was the material of choice; in 1980, SA 516, Grade 70, a plate material
manufactured using fine-grain steel practice, provided a superior combination of toughness and
strength. The latter material, SA 516, Grade 70 is used today for most deaerators and deaerator
storage vessels. Since about 2005, a few deaerator shells and heads have been fabricated from
Type 304L stainless steel in efforts to mitigate CF and FAC corrosion damage. Even small
amounts of chromium in the carbon steel can produce significant immunity to FAC. Please refer
to more information on stainless steels for deaerator shells and components in the following
Discussion section of this report.
Most deaerators purchased by utilities are specified according to the Heat Exchange Institute
(HEI) Standard and Typical Specifications for Tray-Type Deaerators [1]. The standard covers
minimum requirements for design and fabrication Codes (ASME), pressure/vacuum design,
pump section design, nozzle sizes/loads and required accessories. Also covered in the standard
are:
• Minimum corrosion allowance for the deaerator head, shell and nozzles
• Requirements for post weld heat treatment (PWHT; not required by ASME, but considered
necessary to resist in-service CF)
• Nondestructive examination requirements, such as radiographic testing of welds (in addition
to ASME Code)
• Storage tank requirements
• Requirements for accessory equipment such as relief valves, regulating valves/controls and
gauges
2-9
11179101
DA Construction, Mechanical Design, and Materials
Although the Eighth Edition of the HEI Standard was revised in 2008, it does not address some
issues that can improve deaerator longevity. The purchaser of new deaerators should consider the
following additional requirements:
• Require that PWHT of carbon steel vessels be performed at the highest specified temperature
of 1100ºF to 1200ºF (593ºC to 649ºC) in lieu of using ASME approved alternate PWHT
time-temperature cycles (see PWHT in Section 3 of this report).
• Require a baseline WFMT inspection at the fabrication shop to avoid misinterpretation of
weld defects at the first in-service inspection.
• Consider the use of duplex stainless steels in lieu of austenitic stainless steels for improved
resistance to thermal cracking and corrosion fatigue (details below).
• Increasing the tray thickness from 20-gauge to 16-gauge to better withstand upsets and
incidental damage.
Today, all the major deaerator producers for utilities subcontract the fabrication work to outside
tank and pressure vessel shops. Some producers contract to only two or three trusted weld
fabricators, while others contract many shops for fabrication work. Deaerator equipment is
complicated in that many combinations of materials and welding procedures are involved,
including the need for attention to final surface finishes, nondestructive testing, and good control
of PWHT. Quality control becomes more difficult when several fabrication shops are located at
remote distances, or if quality checks are infrequent. All shops holding an ASME stamp must
have a designated quality control (QC) person and QC plan. However, the quality of work in all
ASME shops can vary and in-plant QC programs covering fabrication to ASME requirements
may still be insufficient to ensure the needed quality of deaerators. When purchasing new
deaerator equipment, the utility (end user) should enquire with both the producer and fabricator
as to quality assurance procedures to ensure all requirements are met. Prior to any work being
done it is good practice to visit the fabricator along with the designer/producer to emphasize QC
issues. The end user may also benefit by scheduling his own quality control oversight visits to
the fabricator to emphasize the need for high quality workmanship and to verify that needed
requirements are met.
As noted above, within the past decade or so, a few Type 304L stainless steel deaerator shells
have been manufactured. The intention is to provide better resistance to CF and FAC damage.
There is little doubt that Type 304L stainless steel provides excellent resistance to FAC damage,
but it is questionable as to whether good resistance to CF can be obtained. Austenitic stainless
steels have relatively poor CF resistance and are not commonly used in applications requiring
high corrosion fatigue resistance [2]. Further, it is not practical to post weld heat treat stainless
2-10
11179101
DA Construction, Mechanical Design, and Materials
steel, and residual stresses further reduce the material’s resistance to cyclic fatigue cracking in
water. It should be noted that CF damage occurs more in storage tanks than in deaerators. So far
as is known, no large deaerator storage tanks have been made from stainless steel.
A common corrosion problem with austenitic stainless steel pressure vessels is under-insulation
stress corrosion cracking (UISCC). At least one instance has been reported of under-insulation
stress corrosion cracking of a Type 304L stainless steel deaerator shell [3]. This condition
occurred when chloride contaminated water penetrated the exterior insulation, allowing for
concentration of chlorides and subsequent corrosion cracking. For resistance to UISCC damage,
stainless steel vessels should be painted on exterior surfaces under the insulation.
Duplex stainless steels (DSS) have existed for sixty years. However, the “first generation” of
DSS alloys was not readily welded without loss of toughness and corrosion resistance.
Consequently, their practical use in process applications such as deaerators was severely limited.
Within the last two decades a “second generation” of DSS alloys has come into commercial
production. These grades are characterized by a composition balance, particularly the use of
nitrogen as an alloying element, allowing practical welding without loss of desirable properties.
The most common DSS alloy in use has been Alloy 2205 (UNS S32205), which has corrosion
resistance properties exceeding 316L austenitic stainless steel. However, for use in deaerators,
Alloy 2205 is overqualified—too costly and unnecessarily over-alloyed to simply resist
corrosion from water. More recently, a family of so-called “lean” DSS alloys has been produced
having a better cost advantage for use in mild corrosion-resisting applications such as for
deaerators. Other advantages of DSS (and lean DSS in particular) compared to austenitic alloys
are higher strength, a favorable thermal coefficient of expansion when welded to carbon steel,
and good resistance to UISCC. Not all lean DSS alloys are ASME Code approved for Section
VIII, Division 1 construction. ASME Code Case 2418 qualifies Alloy 2101 (Outokumpu) for
Section VIII construction, and code cases are pending for other manufacturers’ lean DSS alloys.
Deaerator designers and fabricators should be aware that DSS alloys are readily welded, but
must be welded with specific heat input criteria to achieve proper ductility and corrosion
properties. As with any weld fabrication, procedures must be developed and carefully followed
to consistently achieve the intended results. As of the time of this writing, it is unknown as to
whether any deaerator shells or internal components have been built from DSS alloys.
Industrial Heat Recovery Steam Generators have been used to recover heat from a variety of
chemical processes. These units often had stand alone deaerators, and some early versions of
HRSGs used strictly for power generation also used deaerators. However, nearly all the
combined cycle plants built since do not. Rarely there may be a small deaerator section (spray
header and trays) located directly on top of the LP Drum.
2-11
11179101
DA Construction, Mechanical Design, and Materials
In a typical modern combined cycle plant, condensate and makeup water travels from the
condenser, through a heating element close to the stack (often referred to as a feedwater heater or
LP Economizer, then into an LP Drum. The LP drum typically provides head pressure for boiler
feed pumps that feed the IP and HP drums.
Without a standalone deaerator, the HRSG typically relies on deaeration in the hotwell for the
small amounts of makeup water the unit requires. Makeup water often is sprayed high in the
condenser to facilitate the removal of dissolved oxygen in the condenser.
Dissolved oxygen at the condensate pump discharge during operation is typically less than 10
ppb due primarily to the very small amount of makeup required on these units. However, these
units often are called into cycling service. When starting up, the amount of makeup water (and
therefore the amount of dissolved oxygen in the water going to the LP drum) can be much higher
than the desired limit. Corrosion fatigue in the LP drum and associated piping is a significant
corrosion mechanism in some HRSGs.
Reducing agents are not recommended in modern combined cycle plant as they typically contain
no copper alloys in the steam cycle.
2.8 References
2-12
11179101
3
CONFORMANCE TO CODES AND STANDARDS
Several codes and standards are applicable to the safe construction and operation of deaerators.
They provide simple, convenient provisions for the design, fabrication, inspection and repair of
deaerators. A brief discussion of some of these provisions (and limitations) is provided below.
3.1 Codes
A new deaerator should be constructed and stamped in accordance with American Society of
Mechanical Engineers (ASME) Boiler and Pressure Vessel (B&PV) Code Section VIII, Division
1, entitled Rules for the Construction of Pressure Vessels.
3.1.1.1 Design
Design provisions establish minimum wall thicknesses throughout the deaerator based on the
maximum allowable working pressure (MAWP) and the maximum operating temperature that
will be imposed on the deaerator.
The Code designates allowable design stresses for the design of the vessel. They are contained in
the ASME B&PV Code Section II, Part D, entitled Properties. Appropriate safety factors are
incorporated into the allowable design stress values (See Appendix 1, Part D: Basis for
Establishing Stress Values.).
Only approved materials listed in Section II and having maximum use temperatures for Section
VIII of the Code can be used in Code construction. Deaerators are almost always constructed
with carbon steels as the pressure boundary material although many utilities have switched to
low alloy steels with low levels of Chromium to minimize flow accelerated corrosion issues.
Using approved materials establishes a controlled weldability for the material of construction.
3-1
11179101
Conformance to Codes and Standards
3.1.1.4 Fabrication
Provisions include material identification, repair options if defects are present in the materials,
and manufacturing tolerances. Specialized requirements for welded and forged construction are
included in the Code.
3.1.1.5 Welding
Requirements include full penetration, full fusion welding of butt joints and qualification testing
of weld procedures and welders.
1. Post Weld Heat Treatment (PWHT) - Temperature ranges, time and thickness requirements
are delineated. Note: for carbon steels, PWHT is mandatory only for thicknesses over 1½-
inches. For lesser thicknesses, PWHT is optional, usually at the specification of the
purchaser.
2. Inspection - For butt welds, three levels of radiographic inspection (X-ray; RT) of butt welds
are possible—Full (100%), Spot, or No X-ray. RT defines the joint efficiency utilized in the
vessel design with 100%, 85%, and 70% joint efficiency.
The Code requires a hydrostatic test at 1.3 x MAWP (maximum allowable working pressure)
after the completion of the vessel.
The Code requires a U-1 Form to be filled out by the fabricator of the vessel and signed by the
authorized inspector. This provides verification that the vessel was constructed in accordance
with Code requirements. A name plate or impression stamping is affixed on the vessel showing
the manufacturer, MAWP, minimum design temperature, serial number and year built.
In-service boilers and pressure vessels are inspected and repaired according to the rules and
provisions in the NBIC. In many cases, NBIC refers back to the provisions in the ASME original
Code of Construction. The application of the NBIC is determined on a state-by-state basis by
local jurisdictional rules.
ASME and NBIC codes are generic codes that apply to all pressure vessels (and boilers). It is not
possible to include all the special requirements found necessary to ensure the safety of a specific
application, site, product, or application such as a deaerator. For example, except for a specific
provision for establishing a corrosion allowance to be added to the material thickness to allow
3-2
11179101
Conformance to Codes and Standards
for thinning from corrosion, the codes do not contain provisions to mitigate most corrosion
mechanisms. Thus corrosion mechanisms such as stress corrosion cracking, corrosion fatigue,
and hydrogen induced cracking are not addressed nor are they mitigated specifically via Code
provisions.
Some damage mechanisms have been found to be mitigated by PWHT which significantly
reduces the hardness of the weld metal and weld heat affected zone (HAZ), as well as reducing
residual welding stresses. For carbon steels (and certain alloy steels) the ASME Code contains
provisions for reduced PWHT temperatures using extended times at temperatures.
To illustrate, PWHT of carbon steels is usually specified at 1100°F to 1200°F at one hour per
inch of maximum thickness. But the Code allows PWHT at 1050°F to 1150°F at two hours per
inch; 1000°F to 1100°F at four hours per inch; 950°F to 1050°F at ten hours per inch; or 900°F
to 1000°F at twenty hours per inch.
For deaerators, it has been found that cracking of the welds can be mitigated (but not eliminated)
by PWHT. Thus all deaerator vessels should be given a final PWHT irrespective of whether
PWHT is mandatory under Code provisions (see NACE SP0590-2007). Furthermore, it would
seem prudent to also require PWHT at the highest PWHT temperature (1100°F to 1200°F) to
minimize hardness levels and residual stresses in the deaerator.
3.2 Standards
Two standards, specifically applicable to deaerator construction and operation have been
developed. An additional standard, API 579-2/ASME FFS-2, 2009, refers to evaluation
methodology for operation of pressurized equipment with cracking.
NACE SP0590, as the title implies, deals primarily with describing deaerator cracking,
inspection methods to detect that cracking, evaluation methods to assess the significance of the
cracking and possible repair methods.
3.2.2 Heat Exchange Institute, Inc., Standards for Typical Specifications for Tray
Type Deaerators, 8th Edition.
The HEI Standard deals primarily with order requirements for new deaerators which can mitigate
deaerator cracking.
3-3
11179101
Conformance to Codes and Standards
This Standard contains methods for conducting engineering analysis to demonstrate the
structural integrity of an in-service component found to contain a flaw or damage. There are
methodologies specifically prepared for pressurized equipment. The guidelines in this Standard
may be used to make run-repair-replace decisions to help determine if pressurized equipment
containing a flaw can continue to operate safely for some period of time.
The possible application of API 579 to evaluate in-service deaerators found to contain cracking
is discussed in Section 8 of this document.
Many newly manufactured deaerators are built for countries outside of North America, therefore
several of the relevant international regulations and their relevant websites are listed below.
The Pressure Equipment Directive (PED) is the main regulation used, and deaerators could be
built to any Code accepted by the Notified Body. The most relevant codes in Europe are EN
13445 (EU), AD Merkblatt (Germany), CODAP (France), and PD 5500 (UK).
3-4
11179101
Conformance to Codes and Standards
3.2.5.2 Australia
Most of the States will accept the ASME code if the user also agrees with it, but the design will
be verified against AS1210 (Pressure vessels) per State Regulations.
http://infostore.saiglobal.com/store/
OSH Accepts ASME, AS1210, PD 5500 and several other Codes as long as the design is verified
by an approved Design Verification organization and the fabrication is done by and approved
Fabrication Inspection Body.
http://www.osh.dol.govt.nz/
3.2.5.4 Japan
Deaerator construction falls under Ministry of Health Labour and Welfare (MHLW)
Rules/Regulations. Review of the design is performed by an accredited design reviewer. Japan
has its own boiler/pressure vessel standards which ASME usually meets.
3.2.5.5 Malaysia
The Department of Safety and Health (DOSH) accepts the ASME Code, provided fabrication
inspections are performed by an approved (by DOSH) agency and design review is also
performed by that agency or DOSH.
http://www.dosh.gov.my/doshV2/
3.2.5.6 Singapore
The Ministry of Manpower (MOM) accepts the ASME Code with design review and inspection
by an approved agency (by MOM).
http://www.mom.gov.sg/Pages/default.aspx
3.2.5.7 India
Any steam touched vessel or vessel that is a part of the boiler system needs to meet the Indian
Boiler Regulations.
http://dipp.nic.in/boilerrules/index1.htm
http://dipp.nic.in/boiler_rules_updated/contentsregulation.htm
3-5
11179101
11179101
4
DAMAGE MODES IN DA SYSTEMS
Mechanical equipment has a finite life. It can and will deteriorate in service and can fail. Fitness-
for-service (FFS) determination (covered in Section 8 of this document) is a quantitative method
to evaluate the expected rate of deterioration and operating safety of deaerator heater and storage
vessels. The first step in FFS assessment is to identify damage mechanisms or flaw types that are
likely to cause deterioration. In general, the following types of deterioration modes can be
expected in utility deaerator equipment:
• Surface cracking, which may be on the inside or outside
• General loss of metal thickness on the waterside due to corrosion and/or erosion
• Surface pitting on the waterside
• Gradual deterioration of manufacturing flaws by corrosion or cracking
• General metal loss and/or cracking on the outside due to under-insulation corrosion or
general atmospheric corrosion
Knowledge of the damage mechanism(s) present may be obtained by visual examination of the
area of distress in the equipment and comparing the appearance with known exemplar examples
of damage modes. The examination may often be supplemented and verified by performing
metallurgical tests, either by using in-place metallography or by removing samples for laboratory
testing.
Since the failure of several large deaerator storage tanks occurred in the1970s and 1980s many
publications relating to types of vessel damage have been cited. By far the most common types
of damage reported were (1) cracking, and (2) metal loss by erosion-corrosion (now called flow-
accelerated corrosion or FAC). Despite efforts to reduce deaerator and storage tank deterioration
problems, both cracking and metal loss in the vessels are by far the most common problems that
persist today. Yet, many different types of damage, still generally classified as cracking or metal
loss by corrosion, have been reported in conference proceedings, technical journals, and by
deaerator manufacturers. One problem in identifying specific damage mechanisms is that no
systematic studies have yet been specifically directed toward deaerator deterioration and
cracking.
EPRI has published a number of reports on corrosion fatigue both in deaerators and in boiler
tubing [1], [2], [3]. The mechanism is the same whether in the deaerator or boiler, though
thermal stresses are higher in boiler tubing. A comprehensive laboratory research on corrosion
fatigue was conducted regarding the understanding of crack initiation and propagation in boiler
tubes [4], [5].
4-1
11179101
Damage Modes in DA Systems
The following discussions on the various types of damage should help to focus on those most
important to the utility industry. For completeness, the less common types of damage, often more
associated with industrial boilers and deaerator vessels, are also discussed. The various damage
mechanisms are summarized in Table 4.1, and the locations of prevalent damage are shown in
Figure 4.1 and Figure 4.2. General industry terms were used for clarification.
Table 4-1
Deaerator damage mechanisms
4-2
11179101
Damage Modes in DA Systems
Figure 4-1
Damage mechanism locations within the deaerator heater
Figure 4-2
Damage mechanism locations on the inside and outside, including the deaerator storage
tank
4-3
11179101
Damage Modes in DA Systems
4.1.1 General
The consensus of opinion amongst failure analysts in the NACE Committee T7H-7 studying
cracking and rupture of boiler feedwater deaerators is that the main damage mechanism is
corrosion fatigue [6]. This damage mechanism involves formation of protective oxide films
during operation (usually magnetite). The films can rupture if subjected to strain in excess of
0.2% in tension [7]. When an oxide film ruptures, bare metal is exposed to feedwater, and thus a
small anode is formed in the presence of a large cathode (magnetite layer). In the presence of
water, the nascent anode region is immediately converted to more iron oxide. Additional stress
cycles repeat the process until a crack is produced. An important condition for corrosion fatigue
to occur is to have significant and repeated tensile stresses. These stresses have both a mean and
cyclic component. The presence of a high mean stress, such as residual stress from welding,
allows cracking to initiate and propagate with a lower applied tensile stress from operational
loads. Therefore, post weld-heat-treatment (PWHT) may be utilized in new construction to
reduce the vessel’s propensity to CF damage. The cyclic stresses can be from piping vibration,
pressure fluctuations, loads induced by flow (water-hammers), or structural vibratory loads.
4.1.2 Description
Failure analysis studies on deaerators and storage tanks show the following characteristics of
corrosion fatigue [8]:
• Cracks occur in welds and heat affected zones (HAZ).
• Cracks are generally transverse to the welds and HAZ, and occur both parallel and
perpendicular to the hoop stress direction.
• Where cracks occur in base metal, they are opposite external welds or other areas of high
localized stress.
• The worst cracks are located in circumferential and head-to-shell welds in horizontal vessels.
• Cracks are concentrated, but not solely located, below the liquid level in the storage tank.
Cracks are normal to the plate surface and may be very tight.
• Multiple cracks propagate parallel to one another with little branching.
• Cracks are filled with iron oxide corrosion product and only rarely contain trace quantities of
other species such as Al, Si, Ca, Mg, Cl, and S.
• Welds and adjacent base metal, including and extending up to about one inch (25 mm)
beyond the HAZ, have a higher incidence of corrosion pits than the base plate. Pits are often
deeper than they are wide.
• Cracks initiate from corrosion pits.
• Crack tips are blunt.
• Cracks are not limited to any particular material specification.
4-4
11179101
Damage Modes in DA Systems
Most cracking is observed on the waterside of the deaerator storage tank, but cracking in the
deaerator heater is also common. The deepest and greatest number of cracks occur in large
diameter horizontal storage tanks and are generally observed below the water line in the head-to-
shell and circumferential welds. Cracking may be observed adjacent to nozzles, usually on the
waterside but may also be on the outside of nozzles connected to poorly supported piping.
Cracking may be associated on the waterside in regions where attachments or vacuum rings are
welded to the outside surface. Cracking has also been observed on the outside of the vessel
adjacent to saddle welds and similar areas of restraint.
4.1.4 CF Remedies
4.2.1 General
Flow-accelerated corrosion is a form of corrosion that has plagued nuclear and fossil power
plants for many years [10]. (In the United States, flow-accelerated corrosion is commonly known
as erosion-corrosion, where an erosion component is present. However, the term flow-
accelerated corrosion has become more common and is the preferred damage mechanism
term.)The damage mechanism is promoted on the waterside of vessels and piping when the right
combination of liquid flow and environmental conditions exist. FAC damage can occur in many
different metals, but is most prevalent in the carbon steel portion of high temperature piping and
equipment found in power plants. Alloy steels, including low-alloy steel containing chromium,
are resistant to FAC. If the steel can be specified as having a minimum of 0.1% chromium, the
equipment can be expected to have much improved FAC resistance compared to carbon steel
with no chromium [10]. Because water is necessary to remove the oxide layer, FAC does not
occur in equipment or piping transmitting dry or superheated steam.
4-5
11179101
Damage Modes in DA Systems
4.2.2 Description
FAC is almost entirely confined to the shell and nozzle areas of the deaerator heater, as opposed
to the storage tank. Any place where there is a penetration into the deaerator is an area where
FAC can be found. This includes heater drain returns and recirculation lines. In addition, areas
where there is a higher velocity of water and water/steam mixtures are susceptible. This includes
areas under the belly of horizontal spray-tray heaters under the tray stack or around the vortex
breaker on the connections between the deaerator and deaerator storage tank. These areas are
susceptible because of high steam and water flow rates that propel water droplets against the
vessel shell and support components. Problems with water distribution, say from multiple broken
feedwater inlet nozzles, can create an area of FAC where one had not been seen previously. This
is one reason why routine visual inspection of the deaerator and deaerator storage tank is
important.
Other susceptible regions are on the shell at recirculation tees and at the nozzle vortex breaker.
Steam impingement on the inlet target plate of the deaerator is a special case of FAC and is
discussed in the section on steam impingement below.
Unfortunately, some of the conditions that promote FAC in deaerators are defined by the
manufacturer’s design. Horizontal heaters with a relatively small shell-diameter-to-tray-length
ratio will have high velocity steam flows under and through the tray stack, leading to high
impingement conditions. Likewise, recirculation tees, return lines, or nozzles positioned close to
4-6
11179101
Damage Modes in DA Systems
the deaerator shell may promote FAC damage to the shell. Weld overlaying the deaerator shell
with filler metal has been a common method to protect against FAC damage (see Section 9,
Repair Options). Currently, weld overlays using ER70S-B2L consumables have been cited as
best to restore shell thickness and protect against FAC damage [1].
4.3.1 General
Pitting corrosion is a form of localized attack that results in holes in the metal. Pitting may be
described as a cavity with the surface diameter about the same or less than the pit depth. The pit
is most often filled with corrosion product. Pitting in deaerators occurs in the carbon steel and
may be difficult to detect because it is usually covered with layers of oxide. One must remove
the oxide to discover whether pitting has occurred. Oxygen is necessary for pitting to occur in
deaerators. Therefore, deaerators do not experience pitting when operated properly (i.e.,
dissolved oxygen has been removed). Rather, pitting usually occurs when the deaerator system is
shut down for an extended period, allowing oxygen to be absorbed in residual water. Pitting can
occur in all carbon steel areas of the deaerator but is most often found in the storage tank [1].
CF is a damage mechanism that is promoted by pitting [11]. Pitting is not necessary for CF to
occur, but pits serve as potent stress concentrators and reduce the nominal (applied) stress
needed to initiate CF. Pitting is more often experienced in industrial deaerators, but many
instances of utility deaerator pitting have been reported, including the inspection of
approximately 300 utility deaerators/storage tanks, of which 225 were reported to have pitting
corrosion and 170 were cracked [12]. Another report of eighty-four deaerators inspected revealed
that twenty had significant pitting, and cracks were found in nine of them (30% of pitted
deaerator vessels were cracked) [13].
Pitting corrosion may be found in any carbon steel part of the deaerator and storage tank, but is
most prevalent in low regions where water may not drain completely during shutdown or layup
periods. Metal oxide and mineral scale, which is often thicker around welds, can concentrate
impurities and accelerate pitting [14].
As noted above, pitting should not occur during operation if the unit is running properly. Pitting
may be controlled to acceptable levels by maintaining feedwater chemistry within recommended
guidelines and using proper layup procedures [14, 15]. If improper layup occurs, considerable
oxygen pitting damage can occur even if the period of oxygen contamination is short [16].
Oxygen contamination occurs in feedwater systems when mechanical problems exist with
deaerators, feedwater pumps, turbine gland seals, and systems operating under vacuum. Oxygen
4-7
11179101
Damage Modes in DA Systems
intrusion can be caused by improper venting, operational changes that cause cold make-up water,
tray misalignment, malfunction of water spray valves, and operation with an excessive
temperature differential between the deaerator and storage section.
4.4.1 General
Steam Impingement (steam cutting) is a special form of erosion-corrosion, and is closely related
to FAC. Attacked surfaces are locally thinned, usually producing unevenly eroded regions
associated with the steam inlet target to the deaerator. Entrainment of some condensed water is
necessary for impingement attack to occur. As with FAC, affected areas usually show eroded
regions displaying a “metallic” appearance free of oxides, as opposed to iron oxide-covered and
a dull appearance.
The steam target and target support steel are areas of primary metal loss. Associated erosion and
thinning may occur in the deaerator shell adjacent to, and below, the steam inlet target.
Metal loss in the target region is best resolved by constructing the target and supports from Type
304L austenitic, or Type 2205 duplex stainless steel. Positioning of the target plate too close to
the steam inlet nozzle accelerates metal loss on both the target and shell. Mitigation of shell
thinning by steam impingement has been accomplished by welding sacrificial plates over the
affected areas. However, this method involves welding directly to the shell, requiring special
procedures. Subsequent inspection of the shell area under the sacrificial also requires periodic
removal and re-welding of the sacrificial plate.
4.5.1 General
Deaerator vessels and storage tanks are susceptible to mechanical damage. Obvious mechanical
damages are impacts from plant equipment, forklifts, and hoists. These are easily identified and
remediation can be addressed using the NBIC, API 579, or in many cases good engineering
judgments. The most common cause for damage to trays in a spray-tray deaerator is tray upset
when pressure is lost due to turbine load-rejection, control valve failure, or any condition
resulting in flashing of the water in the storage section. Full load trips and loss of turbine
extraction steam may cause flooding of the downcomers, resulting in water being blown upward
against the tray bank. Tray damage can occur if hold-down supports are insufficient or
inadequate.
4-8
11179101
Damage Modes in DA Systems
More subtle forms of mechanical damage occur due to poor design of support structures, water
hammer, or inoperable or poorly maintained expansion plates that result in buckling or high
operational stresses that contribute to other damage mechanisms such as corrosion fatigue.
4.5.2 Identification
Tray position should be verified after flow upsets to check for dislodged trays. Frequent
systematic inspection of the vessels and attached and supporting structures are the best detection
method. Comparison to photographic references is a simple method. More complete surveys
using fixed reference points and comparative to historical trends are also possible if damage is
suspected. One major issue is the ability of the vessels to freely expand and contract. Periodic
inspection and maintenance of the slide plates are very important.
Once mechanical damage is identified, there are two steps for remediation. First, the cause of the
damage must be identified and eliminated. There are various pipe stress analysis programs that
can aid in this evaluation if needed. Generally, a comparison of the current configuration to the
as-built drawings is a good starting point. The second step is to evaluate any secondary damage
such as a contribution to fatigue or SCC and inspect the vessel to detect this damage. The best
protection against damage from tray upsets is to provide operational protection to avoid sudden
pressure loss to the deaerator. Once trays are dislodged, they should be repositioned as soon as
possible to avoid further damage, which can occur even during normal operation.
4.6.1 General
Many other types of damage to deaerators have been reported in technical journals, at
conferences, and in other publications, mostly in non-utility applications. Most of the less
common damage types have been either unsubstantiated by laboratory examination, or were
simply repeated from other publications. These less common damage mechanisms are listed,
with reference to prior sources.
Stress Corrosion Cracking (SCC) has frequently been co-incident with CF damage of carbon
steel deaerator and storage tank shells [6, 17]. The association was cited due to appearance of
crack morphology. Corrosion fatigue is usually characterized by transgranular cracking; cracks
are often wedge shaped or necklaced, with oxides present in the cracks. By contrast, the
morphology of stress corrosion cracking in carbon steel is usually intergranular and branching
with sharp features and tight cracks. Many instances of cracking in carbon steel deaerators and
storage tanks showed mixed intergranular and transgranular cracking, with the cracks branching
toward the tips. Thus, a combination of CF and SCC was suspected. McIntyre [6] reported that
4-9
11179101
Damage Modes in DA Systems
some of the NACE Committee T7H-7D case histories of deaerator cracking were “clearly anodic
stress corrosion cracking due to the carryover of caustic salts from boiler water return.” SCC has
also been attributed to cracking of stainless steel trays and spray valve springs [14]. While both
caustic SCC of carbon steel and chloride SCC of austenitic stainless steel are common damage
modes, they require a concentration mechanism of ions for cracking to occur; thus SCC within
the deaerator is unlikely unless pitting is prevalent. Twigg [18] reported an instance of under-
insulation SCC of a solid 304L stainless steel deaerator heater shell. However, chloride under-
insulation SCC of the exterior of austenitic stainless steel pressure vessels is a well understood
failure mechanism. Chlorides readily concentrate by contaminated water entering through the
insulation and concentrating by flashing on the metal surface. During operation, cracking is
prevalent in regions of high stress such as the weld HAZ.
Thermal fatigue cracking has been cited as a cause for cracking of stainless steel vent piping,
tray enclosures and trays [1]. Thermal stresses result in combined carbon steel-stainless steel
fabrication due a 50% greater coefficient of thermal expansion of stainless compared to carbon
steel. Thermal stresses also result from excessive rates of heating and cooling.
Down-time corrosion can result in general corrosion and pitting on carbon steel surfaces and
should be controlled whenever extended shutdown periods are planned. It is preferred to
maintain a steam blanket on the equipment as this ensures some hot deaerated water for the
subsequent startup [20]. Deaerators and storage tanks should be protected from corrosion during
long periods of down-time or lay-up by blanketing with nitrogen or filling to the vent with water
containing a volatile oxygen scavenger and either ammonia or amine. Units may also be
protected by draining when hot and maintained dry with dehumidified air or desiccant.
Corrosion grooving was described by Twigg [12], where it was observed in storage tanks below
chemical feed inlets.
4-10
11179101
Damage Modes in DA Systems
4.6.7 Galling
Galling damage can be observed on DA systems at the attachment points or on the thermal
expansion plates or pads. Austenitic stainless steel bolts, nuts and studs commonly experience
galling damage
Preferential corrosion can occur along the heat affected zone of welds. Boiler feedwater
preferentially attacks large ferrite grains adjacent to the weld, resulting in ditching [14, 5].
4.6.9 Other
Numerous other damage mechanisms have been cited in deaerator literature, most of which have
not been confirmed by laboratory examination. Those include:
• Stress-Induced Pitting: Localized pitting corrosion that is concentrated in areas of high
residual stress. The higher stress provides energy which accelerates the corrosion process,
thus making the pitting more prone to occur.
• Cavitation: Cavitation has been reported in DA tanks, but it is rare. Cavitation is the pitting
of a metal surface caused by the implosion of vapor bubbles or pockets in a two-phase
system, such as water/steam. Generally the conditions for this to occur in a DA vessel are
rarely seen, unless in an extreme upset condition. Cases reported in the literature are likely
misrepresented as FAC damage.
• CO2 Pitting: Steel vessels can be corrosively attacked by carbonic acid formed from
excessive carbon dioxide in a water-rich system. The attack is a form of undercut pitting. Pits
will generally be deposit free or at best have a small deposit of hematite oxide.
• Chloride Pitting: Chloride pitting is rarely seen in DA tanks unless a severe contamination
event occurs. The pits are severely undercut and energy dispersive analysis of the pit deposit
will show chloride residuals. Stainless steel components in the system can also sustain pitting
or more likely stress corrosion cracking.
• Erosion: Erosion of a DA system is likely FAC or steam impingement, which are discussed
elsewhere in this section.
• Intergranular Corrosion: Intergranular corrosion is generally not seen on DA systems. SCC
can be mistaken for intergranular corrosion.
• Hydrogen Cracking: Hydrogen Cracking of DA systems is usually associated with welding
defects. Poor quality welds, especially those made with consumables that have not been
dried, is the largest cause for hydrogen cracking in welds. In-service hydrogen damage is not
possible given the operational environment of a DA system.
• Caustic Cracking: Caustic cracking is a form of stress corrosion cracking and is discussed in
the above sections.
4-11
11179101
Damage Modes in DA Systems
4.7 References
4-12
11179101
Damage Modes in DA Systems
Copeland, J. F., Eastman, A. D., Schmidt, C. G., “Fatigue an Stress Corrosion Cracking
Evaluations in Deaerators,” Paper #216, proc. Corrosion87, NACE, San Francisco, CA: 1987.
Clevenger, T. G., “Deaerator Cracking – An Industry Update,” Paper #302, proc. Corrosion86,
NACE, Houston, TX: 1986.
4-13
11179101
11179101
5
EFFECTS OF WATER CHEMISTRY ON DAMAGE
INITIATION/PROPAGATION
5.1 Introduction
The feedwater chemistry is critical to controlling and reducing corrosion in the deaerator and
deaerator storage tank. Equally or even more important is the chemistry of any water left in the
deaerator or deaerator storage tank during the time when the unit is off line and the DA is at
atmospheric pressure and exposed to air. Conversely, the proper operation of the deaerator in
removing dissolved oxygen is critical to the control and minimization of corrosion in all
downstream equipment such as feedwater heaters, the economizer, and the boiler.
In utility boilers, the feedwater chemistry program uses only volatile chemicals, typically
ammonia or a neutralizing amine for pH control and a reducing agent such as hydrazine, to
establish a reducing environment for the control of copper alloy corrosion. Reducing agents such
as hydrazine are recommended when copper alloys are present in the feedwater heaters.
Reducing agents are not a substitute for a properly functioning deaerator and should never be
used as such. Although mistakenly referred to as oxygen scavengers, the application of reducing
agents increases the rate of flow-accelerated corrosion in feedwater equipment including the
deaerator and deaerator storage tank.
Nearly all supercritical units in the US, many of the high pressure (> 2600 psig (17.93 MPa))
drum units with full-flow condensate polishing, and all ferrous feedwater heaters operate with
oxygenated treatment (OT). Proper operation of oxygenated treatment provides that once the unit
chemistry is established the deaerator vents are closed, and the deaerator vessel ceases to
function as a deaerator and only serves as a head tank for the boiler feed pumps.
Controlled amounts of oxygen added at the condensate pump discharge minimize flow
accelerated corrosion in the feedwater piping and deaerator during operation. Units on
oxygenated treatment will start up under all volatile treatment guidelines (with no oxygen
addition; also no reducing agent) until the unit is up and running and other chemistry parameters
are close to normal. At this point, the thermal stresses required for corrosion fatigue should have
been relieved, so the addition of oxygen does not increase the propensity of corrosion fatigue in
these units. EPRI recommendations for unit startup and shutdown are that oxygen levels be
maintained below 100 ppb during these transient periods and if possible before firing the boiler
on startups.
5-1
11179101
Effects of Water Chemistry on Damage Initiation/Propagation
Combined cycle power plants generally do not have an independent deaerator. Some HRSGs do
have a small deaerator built into the top of the LP drum. Generally these units depend on the
small amount of makeup water deaeration that occurs in the condenser to maintain the desired
dissolved oxygen limit of < 10 ppb at the condensate pump discharge. In multi-pressure units the
low pressure (LP) evaporator section performs as an efficient deaerator.
This is a very brief overview of feedwater chemistry as it affects the water and steam chemistry
in the deaerator and deaerator storage tank. The EPRI Cycle Chemistry Guidelines for Fossil
Plants [see references on EPRI Chemistry Guidelines at the end of this chapter] provide the
criteria to establish a proper feedwater chemistry program and normal operating limits for their
unit.
The feedwater chemistry of steam cycles operating on All Volatile Treatment and phosphate or
caustic programs are generally the same. The specific limits are dependent on the presence or
absence of copper alloys in the feedwater heaters. If there are copper alloy feedwater heaters, the
pH of the feedwater must be slightly lower to prevent the acceleration of copper corrosion by
excessive amounts of ammonia. The recommended pH limits for mixed metallurgy (copper and
carbon steel in the feedwater section) are 9.0 - 9.3. The pH limits for all ferrous metallurgy (no
copper alloy feedwater heaters) are 9.2 - 9.6.
Ammonia in the feedwater is responsible for maintaining the pH in this range in most fossil-fired
units and HRSGs used primarily for power production. Ammonia addition to the cycle may be
derived from a number of sources, including the direct addition of ammonium hydroxide
solutions to the condensate/feedwater, the breakdown products of neutralizing amines, and the
breakdown product of hydrazine or any other nitrogen-containing reducing agent.
Some ammonia is lost via the deaerator vents, but more leaves with the vacuum pump steam jet
air ejector vent and condensate (if it is not returned to the condenser). Carbon dioxide enters the
feedwater with any makeup water or comes in with air in-leakage. Some carbon dioxide may
also be generated from the degradation of organic (carbon-based) chemical treatments. Carbon
dioxide neutralizes ammonia in the condensate, requiring the addition of more ammonia (or
amine) to bring the pH back up into the control range. The relationship between carbon dioxide
and ammonia concentration can be seen in Figure 5.1. Excessive air in-leakage increases both the
carbon dioxide and ammonia concentrations in the feedwater if the feedwater pH is maintained.
Deaerator design should be capable of reducing dissolved oxygen in the feedwater coming into
the DA to a level of less than 7 ppb. This is generally the design guarantee point. In practice, far
lower levels of dissolved oxygen are achieved in a properly performing deaerator, typically less
than 2 ppb.
Condensate entering the DA should contain less than 10 ppb dissolved oxygen. Higher levels
than 10 ppb typically indicate air in-leakage problems upstream of the condensate pump and
should be addressed at that level. During unit startup condition, large amounts of oxygenated
5-2
11179101
Effects of Water Chemistry on Damage Initiation/Propagation
makeup water are frequently added either directly to the deaerator or indirectly through the
condensate. During part of this time, there may be no steam to the deaerator. Even if steam is
flowing to the DA, the amount of cold makeup water or condensate being fed to the DA is
significant and the quantity of steam is often minimal. Low temperature water can cause the
deaerator to behave more like a water heater. Under these conditions excessive amounts of
oxygen can enter the feedwater.
If there is not a significant decrease in dissolved oxygen across the deaerator (dissolved oxygen
at the deaerator outlet should be less than 7 ppb) and the DA is operating at temperature, it can
be assumed that there are mechanical issues (vents closed, damaged sprays) that are preventing
the deaerator from functioning properly. In this case the deaerator must be taken out of service
and inspected as soon as possible and any mechanical issues corrected.
Corrosion fatigue and flow-accelerated corrosion (both single phase and two-phase) are
functions of the feedwater chemistry. Improving the feedwater chemistry control and startup
chemistry can markedly minimize the conditions contributing to the mechanisms and subsequent
failures.
The phenomenon of FAC is a process in which the normally protective magnetite (Fe3O4) layer
on carbon steel dissolves in a stream of flow water (single phase) or wet steam (two-phase) The
process reduces or eliminates the protective oxide (magnetite) layer and promotes rapid removal
of the base material until the component fails due to excessive wall loss.
When this iron oxide layer is composed nearly exclusively of magnetite (Fe3O4) the porosity of
the magnetite layer and the mobility of the ferrous ion (Fe+2) make it susceptible to removal in
areas of turbulent flow. Chemicals such as hydrazine provide a strongly reducing environment
which promotes the exclusive formation of magnetite. In areas susceptible to FAC, the effect of
the hydrazine is to promote a highly reducing potential which increases the solubitility of
magnetite in solution and accelerates the FAC mechanism.
Small amounts of dissolved oxygen (5-10 ppb) in the water increase the free corrosion potential
by several hundred millivolts and promote the formation of iron oxide hydrate (FeOOH) and
ferric oxide (Fe2O3) particles that form within the porous magnetite layer and inhibit the
diffusion of the Fe2+ ions from the steel surface. In addition, the hematite structure markedly
reduces the solubility of the protective oxide layer, stabilizing it. Increasing the pH of the
feedwater increases the stability of magnetite by increasing the precipitation of the soluble Fe+2
ions, resulting in greater stability of the protective iron oxide layer.
5-3
11179101
Effects of Water Chemistry on Damage Initiation/Propagation
Small amounts of chromium (> 0.3%) in the carbon steel reduce the dissolution rate of the
protective oxide and lower, if not eliminate, the potential for FAC. As iron oxide is removed, the
chromium oxide becomes more concentrated in the passivation layer increasing the stability of
the layer. Studies and experience have shown that even small amounts of chromium make a
significant difference in the rate of FAC in a given area of piping. Typically, when specifying
piping for areas that are susceptible to FAC, 1.25-Cr alloys are favored. When making repairs to
areas of FAC in deaerator shells, chromium-containing weld metal is used for the same reason.
Flow-accelerated corrosion occurs in both single-phase water and in two-phase wet steam (a
mixture of water and steam). In the two-phase fluid the suspended or condensing droplets of
moisture (water) are responsible for the iron oxide dissolution and subsequent FAC mechanism.
The turbulence and motive conditions are typically provided by the rapidly expanding or
collapsing steam. Two-phase FAC is most prominent in areas of pressure or temperature
transients where rapid flashing of superheated liquid or condensing of sub-cooled steam occurs,
resulting in a highly turbulent or high velocity two-phase fluid.
Areas just below feedwater heater drain returns or re-circulating line penetrations are susceptible
to single phase FAC. Excessive flow rates underneath the spray box can cause FAC in an area.
This may be a design issue or can be caused by one or more broken spray nozzles in the
feedwater inlet header. Single phase FAC has a smooth shiny appearance with small divots,
scallops, or chevrons in the metal. Depending on the orientation of flow away from the area, the
chevrons may be more or less circular. The overlapping of the scalloped areas results in a surface
texture has an “orange peel” appearance.
Two-phase FAC is produced when water droplets are entrained in a saturated steam flow and
impinge on a metal surface. They can produce rough patches or very specific areas of corrosion
bordered on both sides by areas where there is no corrosion. This results in a striped or “tiger
striped” appearance. Steam and water mixtures are generally considered more aggressive to
carbon steel than water alone. The area where two-phase corrosion can be found in deaerators is
often at the steam inlet where wet steam may be introduced into the deaerator.
The pH of the feedwater is critical to reducing the rate of FAC in the deaerator. The purpose of a
deaerator is to remove gases from solution by contacting small droplets of the water with steam.
This contact not only removes the dissolved oxygen, but also some ammonia lowering the pH of
the feedwater and increasing the rate of FAC in that area. To counteract this effect, additional
efforts may be required to increase the pH of the feedwater upstream of the deaerator. This will
reduce the potential for FAC in the deaerator.
Unless there are copper-alloy feedwater heaters, EPRI does not recommend the use of any
reducing agent such as hydrazine. Hydrazine has been shown to increase the rate of FAC by
lowering the oxidation-reduction potential and increasing the formation and solubility of ferrous
iron in the protective oxide layer.
5-4
11179101
Effects of Water Chemistry on Damage Initiation/Propagation
When there are copper-alloy feedwater heaters, the plant must balance the potential for copper
corrosion versus the increased risk of FAC. In its most recent guidelines, EPRI recommends
controlling the level of hydrazine such that the oxidation reduction potential (ORP) of the
feedwater is between -300 and -350 mV. To prevent overfeed or underfeed of the reducing agent
in units that cycle or vary load, an on-line ORP or hydrazine analyzer may be needed.
Since deaerators are used in a number of industrial boilers, there are a significant number of
publications in the literature regarding this failure mechanism. Outside the utility industry this
mechanism is also referred to as stress assisted cracking or more generally as environmentally
assisted cracking.
As its name implies, corrosion fatigue is a combination of two separate mechanisms. There is a
corrosion component and a stress-related fatigue component. Stress may be in the form of
residual stresses in the material associated with either formation or welding compounded with
thermal stresses. Severe corrosion fatigue is most often associated with excessive strain (stress)
within the component resulting from restraint of movement of the component, such as
differential thermal expansion restrained by an external attachment, change in component
thickness or component orientation. When the internal strain becomes excessive, a break in the
protective oxide film will result in exposing non-passivated steel to the chemical environment.
Once the oxide layer is cracked, the presence of dissolved oxygen, carbon dioxide, low pH
conditions, and any contamination products or impurities such as chlorides in the water can
attack the underlying metal resulting in the initiation and formation of a pit. The pit induces a
localized stress riser and the site of subsequent attack. The pit growth proceeds with repetitive
stress and corrosion cycles, and cracks develop at the base of the pit in the orientation of the
strain. Dissolved oxygen, low pH conditions, chlorides and other contaminants including
corrosion products migrate to the tips of the cracks continuing the corrosion process and
weakening the metal. Even when the crack becomes filled with iron oxides, the presence of these
oxides can produce stresses within the metal. Repeated stress and corrosion cycles drive the
crack deeper in to the metal. Low pH feedwater, typically driven by the amount of carbon
dioxide or impurities dissolved in the water, as well as conditions of ammonia loss and
phosphate return contributes to the corrosion. The pH at various carbon dioxide and ammonia
concentrations can be seen in Figure 5.2.
Metallographically, corrosion fatigue cracks are transgranular and oxide filled. Areas of
corrosion appear as wide spots in the cracks. These often alternate with straight thin oxide filled
cracks that occur during high stress conditions.
In utility boilers, corrosion fatigue failures more commonly occur in economizers and boiler
tubes, particularly near constraints such as buckstays. In deaerators, corrosion fatigue damage is
primarily located in stressed areas in or near the longitudinal or circumferential welds in the
deaerator and particularly in the deaerator storage tank. Corrosion fatigue may appear as a
pinhole leak or multiple cracks that run axially along a weld or constraint. The mechanical
5-5
11179101
Effects of Water Chemistry on Damage Initiation/Propagation
constraint may be on the inside or outside of the vessel. Regular inspection of the deaerator
welds normally finds corrosion fatigue cracking before it has had a chance to reach deep enough
into the metal to cause a failure.
Removing dissolved oxygen from the feedwater prior to startup reduces the risk of corrosion
fatigue. In Figure 5.3 the Pourbaix diagram on the right shows that the risk of corrosion fatigue is
greater when the oxidation reduction potential is greater than - 0.5 mV with a pH (at 250ºC
(482ºF)) between 6 and 8. (Neutral pH at 250ºC (482ºF) is 5.6) These graphs are at temperatures
higher than typical DA operation, but similar principles apply. In the industry, maintaining
feedwater dissolved oxygen levels below 100 ppb during the entire startup process is considered
good practice.
Achieving low levels of dissolved oxygen during startup often requires purging of condensate
storage tanks with nitrogen, flooding the deaerator with nitrogen prior to filling, or a nitrogen
purge of the deaerator and deaerator storage tank prior to firing. Proper startup technique can
reduce the risk of corrosion fatigue and the prevalence of cracks in the deaerator welds.
Figure 5-1
Change in oxidizing-reducing potential (ORP) and feedwater iron levels (Fe) at the
economizer inlet when Hydrazine (N2H4) is gradually reduced on a 600MW drum rnit with
an all-ferrous feedwater system (Source: R.B. Dooley, J. Mathews, R. Pate and J. Taylor,
“Optimum Chemistry for ‘All-Ferrous’ Feedwater Systems: Why Use an Oxygen
Scavenger,” Proc. 55th International Water Conference, Pittsburgh, PA, Oct. 1–Nov. 2,
1994.)
5-6
11179101
Effects of Water Chemistry on Damage Initiation/Propagation
Figure 5-2
The pH relationship between carbon dioxide and ammonia in feedwater (Source: Cycle
Chemistry Guidelines for Fossil Plants: Phosphate Continuum and Caustic Treatment,
EPRI, Palo Alto, CA: 2004. 1004188)
5-7
11179101
Effects of Water Chemistry on Damage Initiation/Propagation
Figure 5-3
Corrosion fatigue and dissolved oxygen concentrations (Source “Corrosion Fatigue Boiler
Tube Failures in Waterwalls and Economizers Volume 2: Laboratory Corrosion Studies
EPRI, Palo Alto, CA 1992, TR-100455)
5-8
11179101
Effects of Water Chemistry on Damage Initiation/Propagation
1. Cycle Chemistry Guidelines for Fossil Plants: All-Volatile Treatment: Revision 1, EPRI, Palo
Alto, CA: 2002. 1004187.
2. Cycle Chemistry Guidelines for Fossil Plants: Phosphate Continuum and Caustic Treatment,
EPRI Palo Alto, CA 2004 1004188
5-9
11179101
11179101
6
OPERATIONAL FACTORS IN DAMAGE INITIATION
AND PROPAGATION
6.1 General
The objective of this section is to provide guidance to plant engineering and operations as to the
operational factors contributing to deaerator damage. In general, systems that are thermally
stable and operate base loaded are less likely to encounter damage. When upset conditions
(steam hammer, pressure fluctuation, low outlet temperature, high oxygen, and water out of the
vent) become frequent, the root cause(s) of the upsets must be found and corrected. The
following paragraphs cover factors that are believed to contribute to a few of the most serious
damage types in deaerators and deaerator storage tanks.
NACE and others indicate that corrosion fatigue (CF) cracking is most experienced when
frequent load changes occur or the deaerator is alternately subjected to hot steam and cold water
[1, 2, 3]. Operational conditions contributing to water or steam hammer are most likely to
produce severe alternating tensile stresses in the vessel shell, both deaerator and storage tank,
that may lead to fatigue crack initiation and propagation. At the deaerator, steam hammer may be
caused by water entering a steam line or steam-filled space. One case was cited where the water
level in a deaerator was held at an excessively high level during a turbine trip resulting in severe
water hammer and damage to the deaerator supports. Other operational concerns are prolonged
low-load operation, flows beyond design, operation colder than design, and steam temperature in
excess of design. These factors and others that result in repeated pressure surges are most likely
to contribute to CF damage. Copeland [4] reviewed plant operating data and record charts to find
significant stress cycles, especially during start-up. Although deaerator operating pressures are
relatively low, the material thickness in deaerators and storage tanks are such that slightly
increased stresses can be significant, especially when combined with residual stresses inherent in
some non-heat treated vessels. Copeland also monitored operating stresses with strain gages in
vessel areas that were cracked, but did not measure significant stress aberrations.
As noted in Section 4 of this guideline, conditions influencing FAC in deaerators are mostly
defined by the manufacturer’s design, and operating conditions cannot normally be changed to
reduce the problem. An exception is that FAC can occur if the deaerator is operated above design
limits or with damaged spray valves and/or trays, resulting in excessive water impingement on
6-1
11179101
Operational Factors in Damage Initiation and Propagation
lower carbon steel tray enclosure supports. FAC may also be minimized by relocating
recirculation tees or elbows to minimize direct impingement of water onto the carbon steel shell
or attachments.
Corrosion pitting is most likely to occur during shutdown periods and can be minimized by
ensuring the deaerator and storage tank are thoroughly drained and dry. During periods of long
layup the vessels should be protected by using proper layup procedures [2, 5].
Pitting cannot occur during operation if the deaerator is operating normally, but excess oxygen
during operation can be present due to other conditions. One example is high oxygen during non-
steady state operation (startup and shutdown). The length of a stabilization period is system-
specific, but three days of boiler operation to reach steady state may be used as a guideline. If
there is a down-ward trend in oxygen content measurements, steady state condition has not yet
been attained. Other factors contributing to excessive oxygen during operation are improper
design of air inlet, spray valves not installed correctly, water inlet temperature too low and
improper venting.
As noted in Section 4 of this guideline, steam impingement is a special case of FAC. Damage
from steam impingement is confined to the steam inlet region of the deaerator, especially the
target plate and structural attachments. There is little that can be done in operation of the
deaerator to reduce erosion by steam impingement. Rather, the longevity of the target is
dependent on position from the steam inlet (less wear as the distance is increased) and the use of
erosion resistant alloys such as 304L stainless steel for the target plate and attachments.
Many of the operational factors influencing CF also apply to deaerator mechanical damage. As
noted above, full-load rejection from turbine trips may cause flooding of the downcomers
resulting in water being blown upward against the tray bank. Likewise, steam flashing from the
storage tank to the deaerator causes mechanical damage to trays and enclosures, and is related to
excessive pressure drop in the equalizers or excessive pressure drop across the tray bank. This
problem is minimized in the design stage by adequately sizing the equalizers and providing
sufficient height between the bottom of the deaerator and tray stack.
NACE reported in surveys that mechanical damage to trays in utility deaerators was very
common prior to about 1976. Of the 80 installations surveyed, approximately 22% reported
damage to trays or enclosure hardware. Since 1976, the reported incidence of tray damage has
decreased significantly. Damage occurred as a result of plant upsets, especially full-load
rejections. Trays were reported to be dislodged, bent, and broken. Tray end clips, fasteners, and
distribution troughs were also damaged, as well as tray enclosures. More recent design
improvements in tray and hold-down hardware have reduced the incidence of this damage.
6-2
11179101
Operational Factors in Damage Initiation and Propagation
6.7 References
1. Standards and Typical Specifications for Tray Type Deaerators, 8th Ed., Heat Exchange
Institute, Inc., Cleveland, OH: 2008.
2. Jonas, O., Mancini, J., “Corrosion of Deaerators,” ASM Handbook, Vol. 13C, p. 452, 2006.
3. Kelly, J. A., “Operation and Water Chemistry in Deaerator Cracking,” Paper #304, proc.
NACE Corrosion88, NACE, St. Louis, MO: 1988.
4. Copeland, J. F., Eastman, A. D., Schmidt, C. G., “Fatigue an Stress Corrosion Cracking
Evaluations in Deaerators,” Paper #216, proc. Corrosion87, NACE, San Francisco, CA:
1987.
5. “Consensus for the Lay-up of Boilers, Turbines, Turbine Condensers, and Auxiliary
Equipment,” ASME Research Report, CRTD- Vol. 66, 2002.
6-3
11179101
11179101
7
INSPECTION ISSUES/RE-INSPECTION GUIDELINES
7.1 General
The objective of this section is to provide guidance to the plant engineer or corporate team
responsible for ensuring that the deaerator equipment is performing well and evaluating its
physical condition. A written plan for periodic inspection of the deaerator equipment, both
internal and external, is an essential element to accomplish these goals. Fortunately, there is a
wealth of information published to provide insight on what type of damage to look for and the
methods to detect the damage modes. As shown in Section 4 of this report, the most prevalent
damage mechanisms in deaerators occur on the waterside. Therefore, periodic internal
inspections are mandatory. The interval of such inspections varies from unit to unit, and
guidelines to determine inspection intervals are discussed below. Deterioration of deaerators and
storage tanks can also occur on the outside and should not be overlooked in the inspection plan.
Ultimately, the goal of the inspection is to ensure that the equipment is fit for continued
operation until at least the next regularly scheduled shutdown.
The data obtained from the inspection must be reviewed by engineers or inspectors who are
competent to evaluate the results and to determine the equipment serviceability. The data should
also be presented in a form that can be used to assess fitness-for-service (FFS); refer to Section 8
of this document entitled “Fitness-For-Service.” If the inspection reveals conditions that require
repair, the data should be sufficiently complete to assist in decision making in the needed repairs.
Experience in deaerator heater and storage tank failures of the 1980’s focused primarily on
cracking in these vessels. Data by NACE on cracking was presented in the document Standard
Practice – Prevention, Detection and Correction of Deaerator Cracking, 1990 [1]. Updates to
the publication were made in 1996, and again in 2007. The NACE document should be used as
the guide to identify cracking in deaerators and storage tanks. The document covers:
• Personnel qualifications
• NDT equipment
• Weld layout of vessels
• Required surface preparation
• Areas of initial inspection for cracking
• Areas of re-inspection for cracking
7-1
11179101
Inspection Issues/Re-Inspection Guidelines
The last item on this list of cracking, vessel classification, is critical in assessing the interval of
re-inspection and FFS category:
Category I: No relevant discontinuities as defined by the ASME Boiler and Pressure Vessel
Code, Section VIII, Division 1 criteria were detected.
Category II: Discontinuities were detected but weld repairs not required.
The NACE method for categorizing deaerators and tanks is extended in this document to include
damage mechanisms other than cracking, e.g., general metal loss by FAC or steam impingement.
Personal safety for inspectors should be ensured by complying with jurisdictional and plant
procedures for lockout-tagout and safe entry of a confined space. The equipment must be cool
enough and adequately ventilated. Inspectors and nondestructive testing personnel must use the
specified personal protective equipment.
7-2
11179101
Inspection Issues/Re-Inspection Guidelines
3. Ensure inspection personnel are qualified for the required tasks. Walk-down inspections
when the equipment is running should be carried out by plant employees who are trained to
inspect for problem symptoms and operating conditions. Detailed visual and nondestructive
test inspections for internal deterioration, thinning, and vessel cracks should be performed by
properly trained and experienced inspectors. Written NDT procedures should be used. Plant
personnel may be qualified to perform these inspections, including the necessary
nondestructive testing. However, most plants contract with inspection service companies to
do out-of-service deaerator inspections. NDT technicians should be trained and certified in
each inspection method used to inspect deaerators in accordance with the recognized
governing jurisdiction or with either:
4. All inspections except routine daily or weekly walk-down inspections in running mode
by operators should be formally documented. Documentation should be sufficient so that
a person not familiar with the particular inspection can use it to locate all deficiencies and
repeat the inspection and/or test results.
Visually examine the exterior of the heater and storage vessels, looking and listening for steam
or water leakage, moist insulation, hammering in the water inlet piping, water in the vent steam,
and proper steam plume. The NBIC recommends that operators examine for conditions that
should be avoided, including: (a) restricted steam flow, especially with cold feedwater;
(b) operation above maximum capacity; (c) operation below minimum effective capacity; (d)
operating with blended water temperature too high; (e) any significant pressure or temperature
fluctuation; (f) water hammer; and (g) high oxygen in the boiler feedwater from the deaerator.
7-3
11179101
Inspection Issues/Re-Inspection Guidelines
Visual inspection is the most important inspection and detects a majority of problems except
most very tight cracking. Carbon steel component surfaces will usually require additional
cleaning by wire brushing. Almost all problems will be found on the interior of the deaerator
heater and storage tank, and exterior inspection necessitates removal of insulation. Yet, periodic
inspection of the exterior of both the heater and storage tank should be included in the inspection
plan, as cracking can be experienced at supports and unsupported nozzle penetrations. Periodic
exterior inspection that includes removal of insulation should be factored into the overall
inspection plan. Components of the deaerator that should be visually inspected include:
• Vessel Shell: All accessible shell and head surfaces should be inspected for disruption of the
oxide film, looking especially for streaks, pitting, and eroded surfaces (signs of FAC or
steam impingement erosion). Any surface where steam or water enters is a possible region
for deterioration by these damage mechanisms. FAC damage is especially prevalent in the
“underbelly” portion under the tray section, or at the downcomer to the storage tank. Areas of
erosion should be measured for thickness using the UT method. Cracking may be visible,
especially at weld junctures of carbon steel-to-stainless steel, such as the water box or vent
penetrations, but will require additional testing by an appropriate NDT method (discussed
below). Use of low angle illumination parallel to the shell surface increases the effectiveness
of the visual inspection, especially in determination of possible distortion due to weld
overlays.
• Water Distribution System: Water boxes, usually constructed from austenitic stainless steel,
are prone to longitudinal cracking at the juncture to the shell. Repairs are difficult and re-
cracking is almost inevitable. Cracking at the juncture to the shell should be closely
examined and tested for possible crack penetration into the shell. Cracking may also occur at
the vent and spray valve penetrations. The water box stainless steel liner to the shell can be
inspected by removing one or more spray valves. Water distribution header pipes, in lieu of
the older style water boxes, are not rigidly welded to the shell and are normally less affected
by distortion and cracking from differential thermal expansion.
• Vents: The seal welds from vents to the water box may experience cracking, resulting in
water short circuiting the spray valves. Cracking can usually be observed visually, but the
weld should be checked with PT.
• Tray Enclosure: Examine the tray enclosure for distortion, cracking, and signs of water short
circuits.
• Trays: Examine for damage of severely warped trays. Trays should be firmly held in place by
hold-down fixtures. Check for tray displacement and signs of corrosion and cracking.
• Tray Enclosure Supports: Visually inspect welds for cracking, although these components
should also be examined using WFMT or PT.
7-4
11179101
Inspection Issues/Re-Inspection Guidelines
• Steam Baffle: The steam baffle and target plate are subject to deterioration from steam
impingement and can be examined visually. Regions of the vessel head and shell should be
visually examined for local thinning, supplemented by UT thickness inspection.
• Storage Tank: The storage tank is prone to cracking that is usually not detected by visual
inspection. Yet, a 100% visual inspection of the storage tank should be performed looking for
surface discontinuities or visual disturbance of oxides which may indicate local corrosion,
especially at nozzle inlets and outlets. Screen covers on outlet nozzles should be removed for
visual and MT inspection.
As noted above, the reader should consult the NACE Standard [1] for detailed guidelines on
crack detection in deaerator vessels. Supplementary crack test information is provided in this
document:
• Weld Layout: A layout drawing showing all welds must be prepared to ensure accurate
recording during crack detection. It is important to note whether the vessel welds are viewed
from inside or outside if “roll-out” drawings are prepared.
• Surface Preparation: A critical step in the process is surface preparation, which is normally
performed by grit blasting or blend grinding welds for the first inspection, and only power
wire brushing or flap wheel sanding is usually required for re-inspections. A survey by
NACE in 2005 indicated that for internal re-inspections, the majority of respondents used
either abrasive blasting or power brushing. When detected, cracking is usually perpendicular
or longitudinal to the weld. To avoid masking of fine cracks, final grinding should be
performed at a 45 degree angle to the weld cap.
• NDE Equipment: WFMT is almost universally used for crack detection of carbon steel, and
liquid penetrant (PT) is normally used for stainless steel. Both of these methods require good
surface preparation for proper interpretation of surface indications. More recently, users have
found using electromagnetic testing (ET) to be faster and require little or no surface
preparation to detect cracks. The ET method may be used to quickly screen welds for
cracking with suspect crack areas tested more thoroughly with WFMT or PT.
Many insurance and jurisdictional agencies require periodic thickness testing for corrosion
thinning. Thickness testing scans along the shell should never be a substitute for internal visual
inspection. Conditions found by visual inspection should be investigated by UT inspection.
Straight beam UT can provide acceptably accurate results within about +/- 1% when performed
by an experienced technician. Surface preparation by flap wheel sanding is usually required. UT
line scans such as described in API 653 are a way to determine the cross-section thickness profile
for detailed FFS integrity evaluations.
7-5
11179101
Inspection Issues/Re-Inspection Guidelines
7.4.2.4 Reporting:
Accuracy of data recording and reporting is a must and is often overlooked. The function is
simple, yet this mundane task is necessary for the purpose of locating exact test locations in the
future, and to provide guidance for necessary repairs. A complete description of flaws or thin
regions is also necessary should FFS analysis be required. Quality reporting is one reason to use
professional testing services for nondestructive testing; these companies invariably provide
superior report and test documentary services.
Plant or corporate engineering and maintenance personnel should carefully review the inspection
results and, if necessary, perform a more detailed analysis using FFS methods described in
Section 8 of this report. Comparing inspection results from current inspection activities with
those from previous inspections can help to determine if detrimental processes are underway
(e.g. thinning or crack growth) that may compromise the reliability of the deaerator equipment.
The latest inspection results should be the basis for planning the scope and timing of future
inspections. Updated inspection reports for the vessels should be kept in an easily accessible file.
Interval of inspection in this report generally follows the guidelines provided by the NACE
Standard. Closer intervals between inspections may be dictated by insurance or jurisdictional
requirements, which must be followed if they are more stringent than suggestions provided by
NACE or in this report.
Many factors influence deaerator and storage tank deterioration. Included are original design and
fabrication factors, including quality of construction, residual stresses (PWHT), and materials of
construction. Operating conditions must be considered, which include number of pressure cycles,
whether base loaded, and operational upsets. Other factors are past inspection findings, repair
history, and quality of repairs.
Detailed baseline inspection of new vessels should be performed when new, or within three years
of initial operation, to establish an inspection history based on the design, fabrication and
operating factors for that period of operation.
For operating vessels, the NACE Standard recommends inspection intervals not exceeding three
years unless a risk-based FFS approach indicates with confidence that a longer period is
acceptable. Once cracking or wall thinning is found, a more frequent interval (compared to the
most recent inspection interval) should be considered. Vessels classified as Category I in
accordance with the NACE system may be inspected at intervals of up to ten years. In the
Category I condition, annual reviews of operating factors and mechanical/inspection history
should be performed to ensure that changes do not occur that could influence cracking or
corrosion thinning susceptibility. Inspection intervals of vessels in Category II or III should be
performed at one or two-year intervals depending on the severity of deterioration.
7-6
11179101
Inspection Issues/Re-Inspection Guidelines
If a vessel has been found cracked and subsequently repaired, and it has been inspected for two
consecutive intervals with no further cracking (i.e., last two inspections resulted in vessel
Category I), the inspection interval may be extended. For any vessel that is on extended intervals
(not inspected since vessel placed into service – greater than three years; Category I – greater
than ten years; Category II – greater than three years; Category III – greater than one year),
annual reviews of operating factors and mechanical/inspection history shall validate that
operating conditions are consistent with past performance. If operating conditions are
significantly changed or the mechanical/inspection history indicates a change in conditions that
would increase the susceptibility to cracking, consideration should be given to reducing the
corresponding re-inspection interval.
7.6 References
Vormelker, P. R., “Deaerator Inspection and Analysis,” Paper #214, proc. Corrosion87, NACE,
San Francisco, CA: 1987.
Perdomo, J. J., Spry, T. D., “Deaerator Inspection: What to Look For,” Materials Performance,
pp. 52-55, July 2005.
Twigg, R. J., “Reality Check: The Reinspection of Deaerators,” Paper #525, proc. Corrosion 96,
NACE, Houston, TX: 1996.
7-7
11179101
11179101
8
FITNESS-FOR-SERVICE METHODS AND
PROCEDURES
8.1 Introduction
Once the inspection of the deaerator has been completed, the data acquired will need to be
reviewed and assessed. If crack-like flaws, thinning of walls, or significant pitting were found,
then assessment of the damage is required to determine the suitability for continued service as
well as help make run-repair-replace decisions. Both ASME and API, have a common standard
to perform fitness-for-service (FFS) assessments to address the structural integrity of the pressure
containing equipment that was identified through inspection to contain flaws or damage. The
standard is known as API 579-1/ASME FFS-1. Three levels of assessment are provided in the
standard. The Level 1 assessment is the most conservative, but easiest to perform. The Level 2
assessment would require a more detailed evaluation and the Level 3 assessment would require
the most detailed evaluation. Those performing the assessment typically proceed from a Level 1
assessment to Level 3 assessment sequentially until an acceptable result has been determined or
clear course of action is identified. The FFS assessments are recognized by API Codes and
Standards (510, 570, & 653) and by NBIC-23 as suitable means to evaluate the structural
integrity of pressure vessels, piping, and storage tanks. As such, the FFS assessments can be used
for equipment designed to the following ASME B&PV Codes: Section VIII, Division 1; Section
VIII, Division 2; and Section I, as well as API Standards: API 620 and API 650 (these API
standards do not apply to deaerators).
The remainder of this section will provide highlights of the various levels of FFS for typical
damage mechanisms found in deaerators and the basics of the process to assess the current state
of a component’s integrity along with the projected remaining life of the component. For a
detailed “how to’ on performing a FFS assessment, please use the API 579-1/ASME FFS-1
standard with example problems found in API 579-2/ASME FFS-2.
Table 4.1 in Section 4 of this document shows the primary damage mechanisms for DA’s. There
are four primary categories of degradation resulting from the various damage mechanisms: 1)
crack-like flaws, 2) thinning - local and general, 3) pitting, and 4) distortion/deformation. A fifth
category that is not a degradation mechanism for deaerators, but can occur in other vessels, is
brittle fracture. Once a flaw/damage mechanism has been identified, an assessment should be
8-1
11179101
Fitness-For-Service Methods and Procedures
performed to determine suitability for continued operation as well a remaining life assessment.
Table 8.1 shows the FFS assessment procedures for various damage mechanisms that could be
found in deaerators.
The first step in the process is to gather data on the deaerator being evaluated. Some common
data required is the manufacturer’s data report, fabrication drawings, material test reports,
material properties test data, current operating conditions, past inspection reports, and records of
hydro-testing. If some of this information is not available, the analysis accuracy will be
degraded. Assumptions or approximations for unknown data should be conservative.
Deaerators have been manufactured from a variety of materials previously shown in Table 2-1.
Several of these construction materials (SA 212 Grade B, ASTM SA 515 Grade 70, and SA 285
Grade C) have been found to be brittle at room temperature and even higher temperatures,
depending on original grain size as well as head/plate forming temperatures. As such, any
deaerators that were manufactured from these materials as well as all carbon steels and low alloy
steels not listed in the API 579-1/ASME FFS-1 Table 3.2 or ASME Code Section VIII, Division
1, paragraph UCS-66 would need to be evaluated for brittle failure. They may be susceptible to
brittle failures both with and without crack-like flaws. Therefore, an FFS will need to be
performed for brittle fracture first without any flaws and then with any crack-like flaws found
during the inspection. API 579-1/ASME FFS-1 Chapter 3 provides the details needed for the
potential brittle evaluation. The critical exposure temperature (CET) and minimum allowable
temperature (MAT) will need to be determined. For deaerators, the CET will most likely be the
lowest on cold start ups that occur during the winter. The MAT is derived using the API 579-
1/ASME FFS-1 Guideline.
If the inspection of the deaerator reveals areas of the vessel wall thickness that have thinned as a
result of general corrosion or flow-accelerated corrosion (FAC), an FFS assessment per API 579-
1/ASME FFS-1 may need to be performed. If the thinning is within the specified
corrosion/erosion allowance and sufficient thickness is available to account for future loss before
the next scheduled inspection, no FFS assessment is required; otherwise one should be
performed. API 579-1/ASME FFS-1 Chapter 4 provides the details needed for the general metal
loss evaluation.
If the inspection of the deaerator reveals localized areas of the vessel wall thickness that have
thinned as a result of corrosion/erosion, blend grinding that was needed to remove a flaw or
some type of damage, or mechanical damage that created a gouge or dent, an FFS assessment per
API 579-1/ASME FFS-1 may need to be performed. If the thickness is within the specified
corrosion/erosion allowance and sufficient thickness is available to account for future loss before
8-2
11179101
Fitness-For-Service Methods and Procedures
the next scheduled inspection, no FFS is required. However, if this is not the case then an FFS
assessment should be performed. API 579-1/ASME FFS-1 Chapter 5 provides the details needed
for the localized metal loss evaluation. There are basically three-types of localized metal loss:
local thin areas, grooves, and gouges. Local thin areas have a width and length that are the same
order of magnitude, while a groove has a length that is significantly greater than the width. The
gouge has a length that is much greater than the width that and was created by mechanical
means. The flaws must be characterized by analysis before the FFS assessment should be
attempted.
If the deaerator’s inspection reveals pitting corrosion, an FFS assessment per API 579-1/ASME
FFS-1 should be performed. There are four types of pitting categories: widespread pitting,
localized pitting, a combination of widespread pitting and localized metal loss, and a
combination of localized metal loss with pitting confined to the localized metal loss region. API
579-1/ASME FFS-1 Chapter 6 provides the details needed for the pitting corrosion FFS
evaluation.
If the inspection of the deaerator reveals a bulge, out-of-roundness, or general shell distortion, an
FFS assessment per API 579-1/ASME FFS-1 should be performed. API 579-1/ASME FFS-1
Chapter 8 provides the details needed for the shell distortion FFS evaluation.
If the inspection of the deaerator reveals crack-like flaws, an FFS assessment per API 579-
1/ASME FFS-1 should be performed. The crack-like flaw mechanisms for deaerators were
previously shown in Table 4-1 in Section 4. They consist of various types of fatigue (e.g.,
corrosion, thermal, mechanical, and vibratory) and various forms of stress corrosion cracking
(e.g., chloride and caustic). API 579-1/ASME FFS-1 Chapter 9 provides the details needed for
the crack-like flaw FFS evaluation. Stress computation will be required for primary stresses,
secondary stresses, and residual stresses. Material properties (tensile and yield strength) along
with fracture toughness will be required. If the mechanism is fatigue cracking then fatigue data
(da/dn) will be required, and if stress corrosion cracking is the mechanism then crack growth rate
(da/dt) will be required to assess remaining life. Finally, flaw characterization will be required.
For single crack-like flaws, flaw dimensions (length and depth) will be required. For multiple
crack-like flaws (e.g., stress corrosion cracks), evaluation of the interaction of flaws in close
proximity to each other and their idealized shape will be required.
8-3
11179101
Fitness-For-Service Methods and Procedures
If the inspection of the deaerator reveals dents, gouges, or dent/gouge combinations as a result of
inadvertent damage during repair work or inspection, an FFS assessment per API 579-1/ASME
FFS-1 may need to be performed. The effect of the damage can be evaluated for continued
operation or for calculating a reduced maximum allowable operating pressure. For the case of
deaerators, reducing the pressure would not be an option and thus repairs would be required if
under current conditions it is determined not to be suitable for continued operation. API 579-
1/ASME FFS-1 Chapter 12 provides the details needed for the dents and gouges evaluation. A
dent is the inward or outward mechanical deformation of a cross-section of a shell member’s
ideal geometry (e.g., shell bulges inward or outward). A gouge is the mechanical removal of an
elongated section of the wall thickness with the length much greater than the width. The dents
and gouges must be characterized per procedures in Chapter 12.
8-4
11179101
Fitness-For-Service Methods and Procedures
Damage Classes
Part 6 Part 5
Assessment Assessment Part 9
of Pitting of Localized Assessment
Damage Metal Loss of Crack-Like
Flaws
Part 7 Part 8
Assessment of Part 12
Assessment Weld Mis-
of Blisters Assessment of
alignment and
Shell Distortions Dents,
Gouges, and
Dent Gouge
Combinations
Part 9
Assessment
of Crack-Like Part 13
Flaws Assessment
of
Laminations
Part 10
Assessment
Not Applicable of Creep
to Deaerators Damage
Figure 8-1
FFS assessment procedures for various damage classes, taken from Figure 2.1 in API 579-
1/ASME FFS-1 and modified for deaerators.
8-5
11179101
11179101
9
REPAIR OPTIONS/CODE (NBIC) AND INSURANCE
CONSIDERATIONS
Deaerators used by utilities will eventually require repairs, with the type and extent of repair
being dependent on the damage mechanism(s) incurred. This section covers considerations that
are important when damage to the deaerator or deaerator storage vessel must be repaired. The
most common repair is in situ welding.
By far the most costly and time consuming repairs encountered will involve correction of
damage by corrosion fatigue (CF) or flow-accelerated corrosion (FAC). CF damage is most often
repaired by welding within the pressure boundary components of the deaerator and deaerator
storage vessels. FAC damage is most often encountered only in the deaerator, and is of most
concern when the pressure boundary (shell) becomes thinned. Other frequent damage may be
cracking of stainless steel components in the deaerator, especially long seams in the waterbox or
stainless steel vent piping. Such repairs must be carefully performed because the cracking
damage can extend into the pressure boundary of the carbon steel shell. Stainless steel to carbon
steel weld joint repairs are complicated. The largest consideration is the formation of residual
stresses developing from differential thermal expansion coefficients between the alloys. Another
consideration is the dilution of the welds and the local degradation of corrosion resistance
properties.
Several matters are important to review in advance of performing deaerator and storage vessel
repairs:
• Welding on an ASME Code vessel requires repair according to the National Board
Inspection Code (NBIC). The NBIC usually references the welding requirements of the
original construction (most often Section XIII, Div. 1).
• The original ASME U-1 form for the vessel must be used to identify materials of
construction, corrosion allowances, post weld heat treatment (PWHT), and weld joint
efficiencies. The absence of the U-1 form requires extensive metallurgical studies to
characterize the materials and assure the success of the repair. Local code authorities must be
consulted for guidelines in repairing these vessels.
• Previous weld repairs made on the vessel and the materials used, including welding
consumables, must be known.
• Past and current nondestructive testing (NDT) thickness test and crack test results should be
known. It is notable that future damage often originates at locations of past weld repairs
because of the introduction of residual stresses from welding, and sometimes because prior
9-1
11179101
Repair Options/Code (NBIC) and Insurance Considerations
weld repairs were not performed to good standards. Thickness test results are often needed to
decide on the type of repairs needed due to FAC or other corrosion thinning; for example,
whether the repair must be made using a flush insert patch or by weld overlay.
The reader is encouraged to consult the document Repair of Deaerators, #1008069, EPRI, Palo
Alto, CA, 2004 for details on repairs. The 2004 document refers to methods of repair of wasted
areas, selection of filler metals, welding processes used, partial shell replacement/flush patches,
and factors in design, fabrication, PWHT, nondestructive examination, and testing. Additional
commentary not covered in the above EPRI document follows in this document.
Repair by removing cracks or original weld flaws found in the vessel by subsequent
nondestructive testing warrant careful consideration. Many repairs in deaerators and deaerator
storage tanks, especially when the vessels are relatively new, can be made by grinding only,
omitting or deferring weld repairs. This is especially important if the vessel was stress-relieved
(PWHT) during original construction. Owners should not be too quick to make weld repairs
without first performing a Fitness-For-Service (FFS) assessment of the vessel. Refer to Section 8
of this document for details on FFS. The problem with welding on a vessel that received PWHT
during original construction is that field welding methods, even temper-bead welding or welding
followed by PWHT, seldom restore the vessel to the low level of residual stresses that was
achieved in original construction. Temper-bead welding will refine microstructure, but does not
reduce residual stresses from welding. Consequently, after weld repair the vessel may be in a
state more vulnerable to future cracking. Thus, the first step in deciding whether or not to
perform weld repairs is to make a run-repair-replace decision based on the damage mechanism
and extent of damage. This consideration emphasizes the need to perform a base-line
nondestructive test on new deaerator vessels, detecting flaws that resulted from original
fabrication, as opposed to waiting until a later NDT examination and the need to assess whether
a weld flaw is original or a propagating crack.
CF damage almost always occurs at welds. When it is determined that CF damage is present and
that it must be removed, the usual repair method is to grind, re-weld, and re-inspect the weld.
This is frequently very time consuming as the cracked region must be inspected using the liquid
penetrant (PT) or magnetic particle test (WFMT or MT) method to ensure the crack is
completely removed. When cracking is extensive, for example at circumferential weld seams
joining the vessel shell to the head, consideration should be given to completely removing the
head and shell, trimming the damaged regions and re-welding the head to the shell. Finished
welds should be blend ground for good evaluation by NDT. A full circumferential weld seam
may be effectively treated with a PWHT following the weld repair, which may defer or eliminate
future cracking (see also the section following on PWHT).
9-2
11179101
Repair Options/Code (NBIC) and Insurance Considerations
When it is determined that FAC-wasted areas of the deaerator must be repaired, the question is
whether repair should be performed by weld build-up (and/or weld overlay 2) or by installing a
flush (inserted) patch to replace the affected section of the shell. Depending on the vessel size,
most deaerator and deaerator storage tanks have wall thicknesses well under 1.0-inch and
extensive weld build-up is limited due to local distortion of the vessel from welding stresses.
Residual stresses in weld metal are always in tension due to shrinkage of the weld, so the vessel
walls tend to shrink on the inside. Thus, at the risk of excessive distortion, practical weld buildup
or weld overlay may be limited to only a few square feet of the shell.
Weld filler metal for build-up must be selected based on original base metal construction
materials, whereas weld consumables for overlays are best selected with chromium additions to
resist FAC damage. For build-up, the weld metal must meet or exceed the minimum strength and
toughness requirements of the applicable code used for construction of the pressure component.
The weld filler metal used for overlay is less restrictive than for weld build-up since the overlay
thickness does not apply as credit toward ASME minimum allowable wall thickness.
Theoretically, overlays of any composition accepted by ASME Section II, Part C for welding
onto carbon steel may be applied, including low-alloy steels and stainless steels. Studies have
shown that alloys containing chromium (as low as 0.1% chromium) significantly reduce
deterioration by FAC; alloys with 1.25% to 2.25% chromium are very resistant to FAC, and
higher chromium alloy steels are essentially unaffected by FAC [1]. Weld filler metals with
1.25% to 2.25% chromium are crack sensitive and require careful control of preheat and
interpass temperatures to avoid cracking problems. Low alloy carbon steel consumables such as
ER70S-B2L have been used extensively for overlay protection from FAC damage [2]. The
owner should be aware of NBIC restrictions to welding methods that are alternative to PWHT
for Cr-Mo low-alloy welding.
Experience has shown that weld build-up and overlays over large areas are best applied using the
automatic Gas Metal Arc Welding (GMAW) process as opposed to manual welding. However,
not all regions of the deaerator shell around the tray enclosure may be accessible to automatic
welding, and some manual welding is usually needed. With automatic welding, good control is
obtained in preheat temperature, interpass temperature, weld bead placement and bead finish.
Both weld build-up and the finished overlay weld should be tested for welding flaws using
WFMT or MT. It is very important that the quality of weld build-up be free of porosity, slag, and
incomplete fusion when it is part of the pressure boundary shell. Likewise, the weld build-up
quality must be near perfect or the subsequently applied weld overlay will have unacceptable
defects.
As noted above, large or heavy weld build-up can produce unacceptable distortion. In cases
where the shell FAC area is very large or thinning is severe, consideration should be given to
inserting flush patch(es) or replacing a section of the shell. Shell sections are usually replaced
2
Weld build-up is the process of restoring the original shell thickness, whereas weld overlay usually pertains to
weld application of a corrosion or erosion resistant overlay onto the shell. Severely corroded vessels must first be
restored to minimum allowable wall thickness by weld build-up before application of weld overlay.
9-3
11179101
Repair Options/Code (NBIC) and Insurance Considerations
with matching low-carbon pressure vessel steel (refer to the ASME B&PVC Section II
Materials). Some owners have considered replacing shell sections with Cr-Mo low-alloy plate,
such as ASME A387, Grade 11 (1.25 Cr-1/2 Mo), and Grade 12 (1 Cr-1/2 Mo). The Cr-Mo steel
grades should be very resistant to FAC, but owners should be cautioned that weldability tests
must be conducted and precautions taken to ensure the welding procedures are carefully
followed. Further, as noted above, the owner should be aware of NBIC restrictions to welding
methods that are alternative to PWHT for Cr-Mo low-alloys. It is unknown whether installations
with Cr-Mo low-alloy plate have been made.
PWHT is a requirement of original construction for ASME vessels fabricated from P-No. 1
carbon steel when the wall thickness exceeds 1.5 in. (38.1 mm) nominal thickness, including
corrosion allowance (Ref. ASME Section VIII, Div. 1, UCS-56). Since most deaerator and
deaerator storage vessels have lower nominal thicknesses than required by ASME for PWHT,
they are therefore not mandated to receive PWHT during original construction. However, it is
widely accepted by deaerator specifications and standards that PWHT be applied to reduce
susceptibility to CF cracking [3, 4]. Therefore, all new deaerator and deaerator storage vessels
should be purchased in the PWHT condition. Refer also to Section 2 of this document on matters
pertaining to PWHT specifications for deaerators.
Whenever practical, PWHT should be performed in weld repairs by heating the full vessel
circumference. An alternate method is to PWHT the vessel wall, incorporating just the nozzle,
attachment, etc., within the area that is heat treated. However, heat treatment of just a portion of
the wall requires careful control of temperature gradients to reduce residual stresses to a
minimum. Yet, this latter treatment is seldom as effective at reducing residual stresses as heat
treatment of the complete vessel circumference. This is usually a simpler matter in the storage
tank than in the deaerator because of complications when heating dissimilar stainless steel
components within the deaerator.
Use of NBIC alternate post-construction methods (e.g., high preheat and temperbead welding) to
avoid PWHT may be necessary in many cases. While these procedures are widely accepted, they
do not achieve the degree of residual stress relief that is usually obtained in PWHT.
At least one instance is recorded where a large deaerator storage vessel (12.0 ft. (3.66 m)
diameter and 64.0 ft. (19.5 m) long) was PWHT in the field after CF damage was found and
repaired. The vessel was supported on the inside to prevent sagging when heated to 1100ºF
(593ºC) by internal gas firing. No significant CF was found in the vessel during the 25 years
following PWHT.
Insurance companies have great incentive to require that deaerator vessels are inspected and
repaired within best practice guidelines. Further, most insurers (as well as many state and
provincial jurisdictions) may require owners to work within the insurer guidelines for inspection
9-4
11179101
Repair Options/Code (NBIC) and Insurance Considerations
and repair. Likewise, the National Board of Boiler and Pressure Vessel Inspectors (NBIC) have
guidelines covering deaerator repair. These institutions recommend that deaerator owners
maintain a system of deaerator equipment documentation covering data from inspection and
repairs including system design drawings, vessel modification records, histories of inspections
and findings, drawings or sketches showing areas of cracks and repairs, and ASME
Manufacturer’s Data Reports.
Prior to inspection and repair, planning meetings should be held involving the owner, the
insurance inspector (an authorized inspector who holds a valid National Board Commission), and
qualified inspection and repair team representatives. Of particular importance is the insurer’s
attitude toward Fitness-For-Service options available to the user (refer to Section 8) as to
whether the insurer will accept the use of API/ASME FFS-1 procedures to assess the safety of
continued operation of the equipment with known flaws. For example, a recent survey conducted
of both owners and insurance companies indicated that while one insurer accepted the
API/ASME FFS-1 procedures, none of the owners took advantage of this accepted practice. The
same insurer requested that deaerator weld repairs be followed by PWHT, even though the vessel
may not have been PWHT during original construction.
9.6 References
1. Flow-Accelerated Corrosion in Power Plants, TR-106611-R1, Electric Power Research
Institute, Inc., Pleasant Hill, CA: 1998.
2. Repair of Deaerators, EPRI, Palo Alto, CA: 2004. 1008069.
3. Prevention, Detection, and Correction of Deaerator Cracking, NACE SP0590-2007, NACE
International, Houston, TX: 2007.
4. Standards and Typical Specifications for Tray Type Deaerators, 8th Ed., Heat Exchange
Institute, Inc., Cleveland, OH: 2008.
9-5
11179101
11179101
10
CASE STUDIES
A crack was discovered in the center of the vessel on top near a circumferential weld. The crack
was ~twelve inches long on the external surface and over twenty-five inches on internal surface.
The crack initiated on the internal surface at the weld for a vacuum (stiffening ring). The top 70
degrees of the vacuum ring was removed to allow for repair. The weld was removed by grinding,
followed by re-welding.
Figure 10-1
Photograph shows the crack in the circumferential weld of the deaerator storage tank
10-1
11179101
Case Studies
Figure 10-2
Photograph showing the same crack from inside the storage tank
The inlet quill for delivery of water treatment chemicals (sodium sulfite) was improperly
installed—the quill was supposed to penetrate thirty-six inches into the vessel but only extended
two inches. This deficiency did not allow chemicals to dilute, and therefore led to local
erosion/corrosion damage. No leak developed, and the damage was detected via visual
inspection. The repair was made by welding in a patch using NBIC Guidelines and given a local
PWHT.
10-2
11179101
Case Studies
A two-inch diameter feed line made of Type 316L stainless steel was found to be cracked after
three years of service. A leak was noted as an indication of the failure. Metallurgical analysis
identified fatigue as the failure mode. Mechanical analysis noted high vibratory stresses in the
pipe caused by flow turbulence. The pipe was repaired by replacement and reconfiguration of the
supports.
The tank was found cracked during a routine scheduled inspection, which was also the first
inspection since the storage tank was put into service. Because of the extensive nature of the
cracking the tank was replaced.
10-3
11179101
Case Studies
The tank was found cracked during a routine scheduled inspection. Cracking was at the HAZ or
weld toe of a recirculation nozzle attachment and suspected of being caused by inadequate or
poorly adjusted piping supports. The repair chosen was to cut out the old nozzle and
reinforcement pad and install a flush patch. During welding with a 350ºF preheat, cracks
continued to be discovered in the old plate materials adjacent to the new patch. The final solution
was to cut out a larger section to avoid cracking in the old HAZ.
Cracking was found on the longitudinal and circumferential welds during routine inspection.
Metallurgical analysis identified corrosion fatigue, both transverse and longitudinal. Crack
initiation was aided by poor weld workmanship issues, including undercutting and poor fit-up.
Crack propagation was driven by support feet that were not free to move. The repair was done by
local weld repair. The vessel was not PWHT (stress relieved) after the repair. Repairs were
delayed due to the presence of lead paint that had to be remediated.
10-4
11179101
Case Studies
The plant shut down two boilers and laid them up properly with wet layup and nitrogen purge;
however, the two DA systems were neglected. The tanks were inspected as a precursor to re-
commissioning the boilers, and severe internal pitting damage was discovered. The DA tanks
were returned to service, and a three-pound block of zinc was added to each storage tank to arrest
the corrosion. The zinc blocks were dissolved in one month of operation. The tanks were
eventually replaced and the water chemistry was evaluated.
During a plant startup, seals in the steam extraction line failed allowing 500 psi steam to enter
the DA vessel. The vessel catastrophically failed. A low water alarm for the DA occurred
immediately before the failure. The safety valve did not lift and was found to be damaged;
however, it could not be established if the safety valve damage preceded or resulted from the
failure. Steam input would have exceeded the relief valve capacity. Vessel remains were located
400 yards from its original location. Fracture mapping identified the failure origin at the weld
joining the integral DA vessel to the horizontal storage tank. No metallurgical defects were
found.
10-5
11179101
Case Studies
The storage tank, sixty-four feet overall in length and twelve feet in diameter, was found to have
corrosion fatigue cracking at head welds during 1985. Following removal of cracks and re-
welding, the storage tank was isolated from the DA heater and subsequently heat treated in situ.
Before heat treatment it was determined that the original insulation should not be damaged by
the heat treatment temperature of 1100ºF. The inside of the tank was supported to prevent
sagging at the heat treatment temperature. Heating was performed by gas burners inserted
through the manway. Following the heat treatment cycle the insulation was found undamaged.
Periodic WFMT inspections since 1985 have not revealed cracking. The tank is still in service;
current WFMT inspections are performed at three-year intervals.
An original tray-spray deaerator heater was installed in 1980 but was replaced due to
performance problem with a vertical spray-scrubber heater in 1983. The replacement DA was not
post weld heat treated. During the twenty-six years of service, the vessel exhibited corrosion
fatigue cracking at welds and significant thinning from flow accelerated corrosion of the shell
adjacent to baffle plates (Figure 10.3). An inspection in 2009 revealed thinning at a weld joint,
and a region of incomplete penetration of the weld (Figure 10.4). The unit was replaced with a
new vertical spray-scrubber during 2009.
10-6
11179101
Case Studies
Figure 10-3
Flow accelerated corrosion on the ID of the spray-scrubber deaerator. The circumferential
weld was previously ground.
Figure 10-4
Flow accelerated corrosion of the DA heater circumferential weld seam revealed a partial
penetration weld. The DA heater was replaced.
10-7
11179101
Case Studies
Two original carbon steel deaerator heaters were replaced during 2005 due to problems with
severe flow accelerated corrosion. The new DA shells were fabricated from Type 304L stainless
steel—carbon steel tray support channels were welded to the shell. The new units operated
satisfactorily for approximately three years when a through-wall crack was discovered in the
proximity of a support saddle attachment on one unit. Internal examination revealed cracking
adjacent to an internal tray enclosure support channel. Further examination revealed that the
carbon steel support channel had been rigidly welded to both sides of the stainless vessel walls
(Figure 10.5). The analysis concluded that the difference in coefficient of thermal expansion for
the stainless steel shell and carbon steel supports produced high stresses and yielding of the
vessel wall, ultimately resulting in cracking. A check of the original vessel design revealed that
the support channel was to be welded to one side only, and the channel was to “float” on the
opposing wall. Additionally, the proximity of the support to the external saddle attachment
increased rigidity and local stresses. A check of the mating stainless steel heater showed cracking
in the same location, but the crack had not progressed completely through the vessel wall.
Figure 10-5
Drawing shows the failure location in the deaerator heater
10-8
11179101
11179101
Export Control Restrictions The Electric Power Research Institute Inc.,
Access to and use of EPRI Intellectual Property is granted (EPRI, www.epri.com) conducts research and
with the specific understanding and requirement that development relating to the generation, delivery
responsibility for ensuring full compliance with all applicable and use of electricity for the benefit of the public.
U.S. and foreign export laws and regulations is being An independent, nonprofit organization, EPRI
undertaken by you and your company. This includes an
brings together its scientists and engineers as well
obligation to ensure that any individual receiving access
hereunder who is not a U.S. citizen or permanent U.S. as experts from academia and industry to help
resident is permitted access under applicable U.S. and address challenges in electricity, including
foreign export laws and regulations. In the event you are reliability, efficiency, health, safety and the
uncertain whether you or your company may lawfully obtain environment. EPRI also provides technology, policy
access to this EPRI Intellectual Property, you acknowledge and economic analyses to drive long-range
that it is your obligation to consult with your company’s legal
research and development planning, and supports
counsel to determine whether this access is lawful.
Although EPRI may make available on a case-by-case research in emerging technologies. EPRI’s
basis an informal assessment of the applicable U.S. export members represent more than 90 percent of the
classification for specific EPRI Intellectual Property, you and electricity generated and delivered in the United
your company acknowledge that this assessment is solely States, and international participation extends to 40
for informational purposes and not for reliance purposes. countries. EPRI’s principal offices and laboratories
You and your company acknowledge that it is still the
are located in Palo Alto, Calif.; Charlotte, N.C.;
obligation of you and your company to make your own
assessment of the applicable U.S. export classification and Knoxville, Tenn.; and Lenox, Mass.
ensure compliance accordingly. You and your company
understand and acknowledge your obligations to make a Together…Shaping the Future of Electricity
prompt report to EPRI and the appropriate authorities
regarding any access to or use of EPRI Intellectual Property
hereunder that may be in violation of applicable U.S. or
foreign export laws or regulations.
© 2010 Electric Power Research Institute (EPRI), Inc. All rights reserved.
Electric Power Research Institute, EPRI, and TOGETHER…SHAPING THE
FUTURE OF ELECTRICITY are registered service marks of the Electric
Power Research Institute, Inc.
1021199
11179101